Add listing of all fwd and bwd device ops and instances.

BACKWARD_CONVOLUTION_DEVICE_OPS_SUMMARY.md

@@ -0,0 +1,72 @@
+# Backward Convolution Device Operations - Comprehensive Summary
+
+**Generated:** October 29, 2025  
+**Namespace:** `ck::tensor_operation::device`  
+**Location:** `/library/include` and `/library/src` directories
+
+---
+
+## Overview
+
+This document provides a comprehensive list of all device operations used for backward convolutions (both weight gradients and data gradients) in the Composable Kernel library.
+
+### Total Statistics
+- **Unique Device Operation Types:** 15
+- **Backward Weight Operations:** 9 types (885 instantiations)
+- **Backward Data Operations:** 7 types (1046 instantiations)
+- **Total Template Instantiations:** 1827
+
+---
+
+## Backward Weight Device Operations
+
+| # | Device Operation Name | Instantiations | Files | Primary Location |
+|---|----------------------|----------------|-------|------------------|
+| 1 | `DeviceGroupedConvBwdWeight` | 342 | 157 | `grouped_convolution_backward_weight.hpp` |
+| 2 | `DeviceGroupedConvBwdWeightMultipleD_Xdl_CShuffle` | 161 | 2 | `device_grouped_conv_bwd_weight_xdl_bilinear_instance.hpp` |
+| 3 | `DeviceGroupedConvBwdWeight_Xdl_CShuffle` | 108 | 1 | `device_grouped_conv_bwd_weight_xdl_instance.hpp` |
+| 4 | `DeviceConv2dBwdDataXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K` | 104 | 4 | `device_conv2d_bwd_data_xdl_nhwc_kyxc_nhwk_bf16_instance.cpp` |
+| 5 | `DeviceGroupedConvBwdWeightTwoStage_Xdl_CShuffle` | 96 | 1 | `device_grouped_conv_bwd_weight_two_stage_xdl_instance.hpp` |
+| 6 | `DeviceGroupedConvBwdWeight_Wmma_CShuffle` | 41 | 1 | `device_grouped_conv_bwd_weight_wmma_instance.hpp` |
+| 7 | `DeviceGroupedConvBwdWeight_Xdl_CShuffleV3` | 16 | 1 | `device_grouped_conv_bwd_weight_v3_xdl_instance.hpp` |
+| 8 | `DeviceGroupedConvBwdWeightMultipleD` | 14 | 2 | `grouped_convolution_backward_weight_bilinear.hpp` |
+| 9 | `DeviceGroupedConvBwdWeight_Dl` | 3 | 1 | `device_grouped_conv_bwd_weight_dl_instance.hpp` |
+| **Total** | **Backward Weight Operations** | **885** | **170** | |
+
+---
+
+## Backward Data Device Operations
+
+| # | Device Operation Name | Instantiations | Files | Primary Location |
+|---|----------------------|----------------|-------|------------------|
+| 1 | `DeviceGroupedConvBwdDataMultipleD_Xdl_CShuffle_v1` | 370 | 4 | `device_grouped_conv_bwd_data_transpose_xdl_instance.hpp` |
+| 2 | `DeviceConvNdBwdDataNwcKxcNwk_Xdl` | 312 | 12 | `device_conv1d_bwd_data_xdl_nwc_kxc_nwk_bf16_instance.cpp` |
+| 3 | `DeviceGroupedConvBwdDataMultipleD` | 156 | 76 | `grouped_convolution_backward_data.hpp` |
+| 4 | `DeviceConv2dBwdDataXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K` | 104 | 4 | `device_conv2d_bwd_data_xdl_nhwc_kyxc_nhwk_bf16_instance.cpp` |
+| 5 | `DeviceGroupedConvBwdDataMultipleD_Wmma_CShuffle` | 66 | 2 | `device_grouped_conv_bwd_data_wmma_f16_instance.hpp` |
+| 6 | `DeviceConvBwdData` | 32 | 16 | `convolution_backward_data.hpp` |
+| 7 | `DeviceConvNdBwdDataNwcKxcNwk_Dl` | 6 | 3 | `device_conv2d_bwd_data_dl_nhwc_kyxc_nhwk_f16_instance.cpp` |
+| **Total** | **Backward Data Operations** | **1046** | **117** | |
+
+### Grand Total: 1827 Template Instantiations across 15 Device Operation Types
+
+---
+
+## Output Files
+
+### Complete Instantiation Files
+
+1. **`backward_conv_all_instantiations.txt`**
+   - COMPLETE listing of ALL template instantiations
+   - Human-readable format with line numbers
+   - Organized by operation type, then by file
+
+2. **`backward_conv_all_instantiations.json`**
+   - Structured JSON for programmatic instantiation generation
+   - Separated into `backward_weight_operations` and `backward_data_operations`
+   - Each instantiation includes full text and parsed parameters
+   - Ready for automated code generation
+
+3. **`BACKWARD_CONVOLUTION_DEVICE_OPS_SUMMARY.md`** (This file)
+   - Executive summary with tables
+   - Quick reference for all backward operations

FORWARD_CONVOLUTION_DEVICE_OPS_SUMMARY.md

@@ -0,0 +1,377 @@
+# Forward Convolution Device Operations - Comprehensive Summary
+
+**Generated:** October 29, 2025  
+**Namespace:** `ck::tensor_operation::device`  
+**Location:** `/library/include` and `/library/src` directories
+
+---
+
+## Overview
+
+This document provides a comprehensive list of all device operations used for forward convolutions in the Composable Kernel library, along with their template instantiations.
+
+### Total Statistics
+- **Unique Device Operation Types:** 8
+- **Total Template Instantiations:** 1,000
+- **Files Analyzed:** 420
+- **Header Files:** Located in `/library/include/ck/library/tensor_operation_instance/gpu/`
+- **Source Files:** Located in `/library/src/tensor_operation_instance/gpu/`
+
+---
+
+## Device Operations Table
+
+### Grouped Convolution Device Operations
+
+| # | Device Operation Name | Instantiations | Files | Description |
+|---|----------------------|----------------|-------|-------------|
+| 1 | `DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle` | 570 | 9 | XDL-based with multiple ABD inputs, CShuffle optimization |
+| 2 | `DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3` | 141 | 2 | Version 3 with advanced pipeline scheduling |
+| 3 | `DeviceGroupedConvFwdMultipleD_Wmma_CShuffle` | 42 | 1 | WMMA-based (Wave Matrix Multiply Accumulate) |
+| 4 | `DeviceGroupedConvFwdMultipleD_Xdl_CShuffle_Large_Tensor` | 12 | 1 | Optimized for large tensor dimensions |
+| 5 | `DeviceGroupedConvFwdDlMultipleD_NHWC_KYXC_NHWK` | 6 | 1 | Direct Load implementation |
+| **Total** | **Grouped Convolution Operations** | **771** | **14** | |
+
+### Non-Grouped Convolution Device Operations
+
+| # | Device Operation Name | Instantiations | Files | Description |
+|---|----------------------|----------------|-------|-------------|
+| 1 | `DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K` | 156 | 4 | Standard 2D convolution, explicit layout |
+| 2 | `DeviceConv2dFwdXdl_C_Shuffle_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K` | 61 | 1 | 2D convolution with C-Shuffle optimization |
+| 3 | `DeviceConvFwd` | 12 | 6 | Generic interface type |
+| **Total** | **Non-Grouped Operations** | **229** | **11** | |
+
+### Grand Total: 1,000 Template Instantiations across 8 Device Operation Types
+
+---
+
+## Device Operations List
+
+### 1. DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle
+**Most commonly used device operation for forward convolutions**
+- **Template Instantiations:** 570
+- **Files:** 9 header files
+- **Primary Header:** `grouped_conv_fwd/device_grouped_conv_fwd_xdl_instance.hpp`
+- **Description:** XDL-based grouped convolution with multiple auxiliary inputs/outputs and CShuffle optimization
+- **Variants:**
+  - Standard instances (BF16, F16, F32, INT8, F8, BF8)
+  - 16x16 MFMA instances
+  - NCHW layout instances
+  - Generic instances
+  - Compute-friendly instances with FP8
+
+**Key Files:**
+- `device_grouped_conv_fwd_xdl_instance.hpp` - Main instantiations (193 instances)
+- `device_grouped_conv_fwd_xdl_bilinear_instance.hpp` - With bilinear fusion (80 instances)
+- `device_grouped_conv_fwd_xdl_scale_instance.hpp` - With scale operation (80 instances)
+- `device_grouped_conv_fwd_xdl_dynamic_op_instance.hpp` - With dynamic operations (64 instances)
+- `device_grouped_conv_fwd_xdl_outelementop_instance.hpp` - With custom output ops (80 instances)
+- `device_grouped_conv_fwd_xdl_merged_groups_instance.hpp` - Merged groups optimization (21 instances)
+- `device_grouped_conv_fwd_xdl_scaleadd_ab_instance.hpp` - ScaleAdd operations (20 instances)
+- `device_grouped_conv_fwd_xdl_scaleadd_scaleadd_relu_instance.hpp` - Fused operations (16 instances)
+- `device_grouped_conv_fwd_xdl_binary_outelementop_instance.hpp` - Binary output ops (16 instances)
+
+---
+
+### 2. DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K
+**Standard 2D convolution forward operation**
+- **Template Instantiations:** 156
+- **Files:** 4 source files
+- **Description:** XDL-based 2D convolution with explicit NHWC layout
+
+**Instantiation Files:**
+- `conv2d_fwd/device_conv2d_fwd_xdl_nhwc_kyxc_nhwk_bf16_instance.cpp` (39 instances)
+- `conv2d_fwd/device_conv2d_fwd_xdl_nhwc_kyxc_nhwk_f16_instance.cpp` (39 instances)
+- `conv2d_fwd/device_conv2d_fwd_xdl_nhwc_kyxc_nhwk_f32_instance.cpp` (39 instances)
+- `conv2d_fwd/device_conv2d_fwd_xdl_nhwc_kyxc_nhwk_int8_instance.cpp` (39 instances)
+
+---
+
+### 3. DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3
+**Version 3 of XDL CShuffle implementation**
+- **Template Instantiations:** 141
+- **Files:** 2 header files
+- **Description:** Advanced XDL implementation with compute-friendly scheduling and memory optimizations
+
+**Key Files:**
+- `device_grouped_conv_fwd_xdl_comp_instance.hpp` (49 instances) - Compute-optimized variants
+- `device_grouped_conv_fwd_xdl_mem_instance.hpp` (92 instances) - Memory-optimized variants
+
+**Features:**
+- BlockGemmPipelineScheduler variants (Interwave, Intrawave)
+- Multiple pipeline versions (v1, v3, v4, v5)
+- Support for BF16, F16, F32, TF32, INT8
+
+---
+
+### 4. DeviceConv2dFwdXdl_C_Shuffle_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K
+**C-Shuffle variant for 2D convolution**
+- **Template Instantiations:** 61
+- **Files:** 1 source file
+- **File:** `conv2d_fwd/device_conv2d_fwd_xdl_c_shuffle_nhwc_kyxc_nhwk_f16_instance.cpp`
+- **Data Type:** F16 only
+- **Description:** CShuffle optimization for channel dimension
+
+---
+
+### 5. DeviceGroupedConvFwdMultipleD_Wmma_CShuffle
+**WMMA-based grouped convolution**
+- **Template Instantiations:** 42
+- **Files:** 1 header file
+- **File:** `grouped_conv_fwd/device_grouped_conv_fwd_wmma_instance.hpp`
+- **Description:** Uses Wave Matrix Multiply Accumulate (WMMA) instructions
+- **Data Types:** F16, INT8
+- **Block Sizes:** 32, 64, 128, 256
+
+---
+
+### 6. DeviceConvFwd
+**Generic convolution forward interface**
+- **Template Instantiations:** 12
+- **Files:** 6 files (headers and sources)
+- **Description:** High-level interface type for convolution forward operations
+- **Usage:** Factory pattern and API definitions
+
+---
+
+### 7. DeviceGroupedConvFwdMultipleD_Xdl_CShuffle_Large_Tensor
+**Large tensor optimization**
+- **Template Instantiations:** 12
+- **Files:** 1 header file
+- **File:** `grouped_conv_fwd/device_grouped_conv_fwd_xdl_large_tensor_instance.hpp`
+- **Description:** Optimized for large tensor dimensions
+- **Data Types:** BF16, F16, F32, TF32, INT8
+
+---
+
+### 8. DeviceGroupedConvFwdDlMultipleD_NHWC_KYXC_NHWK
+**Direct Load (DL) implementation**
+- **Template Instantiations:** 6
+- **Files:** 1 header file
+- **File:** `grouped_conv_fwd/device_grouped_conv_fwd_dl_instance.hpp`
+- **Description:** Direct load convolution implementation for specific layouts
+- **Data Types:** F16, F32
+
+---
+
+## File Organization
+
+### Header Files (`/library/include/ck/library/tensor_operation_instance/gpu/`)
+
+#### Grouped Convolution Forward (`grouped_conv_fwd/`)
+1. `device_grouped_conv_fwd_dl_instance.hpp` - Direct Load implementations
+2. `device_grouped_conv_fwd_wmma_instance.hpp` - WMMA implementations
+3. `device_grouped_conv_fwd_xdl_instance.hpp` - **Main XDL implementations**
+4. `device_grouped_conv_fwd_xdl_bilinear_instance.hpp` - With bilinear fusion
+5. `device_grouped_conv_fwd_xdl_binary_outelementop_instance.hpp` - Binary output ops
+6. `device_grouped_conv_fwd_xdl_comp_instance.hpp` - Compute-optimized
+7. `device_grouped_conv_fwd_xdl_dynamic_op_instance.hpp` - Dynamic operations
+8. `device_grouped_conv_fwd_xdl_large_tensor_instance.hpp` - Large tensor optimization
+9. `device_grouped_conv_fwd_xdl_mem_instance.hpp` - Memory-optimized
+10. `device_grouped_conv_fwd_xdl_merged_groups_instance.hpp` - Merged groups
+11. `device_grouped_conv_fwd_xdl_outelementop_instance.hpp` - Custom output operations
+12. `device_grouped_conv_fwd_xdl_scale_instance.hpp` - With scale operation
+13. `device_grouped_conv_fwd_xdl_scaleadd_ab_instance.hpp` - ScaleAdd on inputs
+14. `device_grouped_conv_fwd_xdl_scaleadd_scaleadd_relu_instance.hpp` - Fused operations
+
+### Source Files (`/library/src/tensor_operation_instance/gpu/`)
+
+#### Conv2D Forward (`conv2d_fwd/`)
+- Standard XDL instances for BF16, F16, F32, INT8
+- C-Shuffle variant for F16
+
+#### Grouped Conv1D/2D/3D Forward
+Organized by:
+- **Algorithm:** `dl/`, `wmma/`, `xdl/`
+- **Optimization:** `comp/`, `mem/`, `large_tensor/`, `merged_groups/`
+- **Data Type:** Per file (bf16, f16, f32, int8, fp8, bf8)
+- **Layout:** Encoded in filename (nhwgc, ngchw, etc.)
+
+---
+
+## Template Instantiation Patterns
+
+### Common Parameters
+Template instantiations typically include:
+- **Spatial Dimensions:** 1D, 2D, 3D (NDimSpatial)
+- **Layouts:** NHWC, NCHW, and grouped variants (NHWGC, NGCHW, etc.)
+- **Data Types:** BF16, F16, F32, TF32, INT8, F8, BF8
+- **Accumulator Type:** F32, INT32
+- **Block Sizes:** 32, 64, 128, 256
+- **Thread Tile Sizes:** MPerBlock, NPerBlock, KPerBlock
+- **MFMA Sizes:** 16x16, 32x32
+- **Pipeline Stages:** 1-2 stages
+- **Element-wise Operations:** PassThrough, Scale, ScaleAdd, Bilinear, ReLU, Clamp, etc.
+
+### Example Instantiation (from DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle)
+```cpp
+DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<
+    NDimSpatial,    // 2 or 3
+    ALayout,        // NHWGC, NGCHW, etc.
+    BLayout,        // GKYXC, GKCYX, etc.
+    DsLayout,       // Additional inputs layout
+    ELayout,        // Output layout
+    BF16,           // Input data type
+    BF16,           // Weight data type
+    F32,            // Accumulator type
+    BF16,           // CShuffle data type
+    DsDataTypes,    // Additional input types
+    BF16,           // Output data type
+    PassThrough,    // Input element-wise op
+    PassThrough,    // Weight element-wise op
+    OutElementOp,   // Output element-wise op
+    ConvSpec,       // Convolution specialization
+    GemmMNKPadding, // GEMM specialization
+    1,              // NumGemmKPrefetchStage
+    256,            // BlockSize
+    128,            // MPerBlock
+    128,            // NPerBlock
+    32,             // KPerBlock
+    8,              // AK1
+    8,              // BK1
+    32,             // MPerXdl
+    32,             // NPerXdl
+    2,              // MXdlPerWave
+    2,              // NXdlPerWave
+    // ... additional block transfer parameters
+>
+```
+
+---
+
+## Key Findings
+
+1. **DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle is the primary device operation** with 570 instantiations across 9 different header files for various use cases and fusion patterns.
+
+2. **Eight distinct device operation types** are used for forward convolutions, each optimized for different scenarios:
+   - XDL (Matrix Core) based operations
+   - WMMA (Wave Matrix Multiply) based operations  
+   - Direct Load (DL) based operations
+   - Specialized variants for large tensors, merged groups, and memory optimization
+
+3. **Multiple optimization strategies:**
+   - Compute-friendly scheduling (Interwave, Intrawave)
+   - Memory access patterns (Inter-wave, Intra-wave)
+   - Pipeline versions (v1, v3, v4, v5)
+   - CShuffle for efficient data movement
+
+4. **Comprehensive data type support:** BF16, F16, F32, TF32, INT8, F8, BF8
+
+5. **Extensive fusion support:** Operations can be fused with Scale, ScaleAdd, Bilinear, ReLU, Clamp, and combinations thereof.
+
+---
+
+## Output Files
+
+### Complete Instantiation Files (Recommended)
+
+1. **`forward_conv_all_instantiations.txt`** (593KB, 4,271 lines)
+   - **COMPLETE listing of ALL 1,000 template instantiations**
+   - Human-readable format with line numbers
+   - Every instantiation shown in full detail
+   - Organized by device operation, then by file
+
+2. **`forward_conv_all_instantiations.json`** (1.4MB)
+   - **Structured JSON for programmatic instantiation generation**
+   - Complete instantiation text for each template
+   - Includes parsed parameters (data types, block sizes)
+   - Hierarchical organization: device_operation → file → instantiations[]
+   - Ready for automated code generation tools
+
+### Summary and Quick Reference Files
+
+3. **`FORWARD_CONVOLUTION_DEVICE_OPS_SUMMARY.md`** (This file)
+   - Executive summary with tables
+   - Device operation descriptions
+   - File organization reference
+
+4. **`forward_convolution_device_ops_report.txt`** (33KB)
+   - High-level summary of all device operations
+   - Lists all files containing each device operation
+   - Quick reference guide
+
+5. **`forward_conv_device_ops_detailed_report.txt`** (43KB)
+   - Detailed report with sample template instantiations
+   - Shows first 3 examples per file
+   - Includes line numbers and file locations
+
+6. **`forward_convolution_device_ops_data.json`** (34KB)
+   - Basic JSON format with summary data
+   - File paths and instantiation counts
+
+---
+
+## Usage Examples
+
+### Example 1: DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle with BF16
+Found in: `device_grouped_conv_fwd_xdl_instance.hpp`
+
+This is the most versatile device operation with:
+- Generic instances for all block sizes
+- Optimized instances for small conv.K and conv.C
+- NCHW layout support
+- 16x16 MFMA support
+- FP8 compute support
+
+### Example 2: DeviceGroupedConvFwdDlMultipleD_NHWC_KYXC_NHWK
+Found in: `device_grouped_conv_fwd_dl_instance.hpp`
+
+Direct Load implementation with:
+- F16 and F32 support
+- Specific block configurations
+- NHWC layout only
+
+### Example 3: DeviceGroupedConvFwdMultipleD_Wmma_CShuffle  
+Found in: `device_grouped_conv_fwd_wmma_instance.hpp`
+
+WMMA-based implementation with:
+- F16 and INT8 support
+- Multiple block sizes (32, 64, 128, 256)
+- 16x16 WMMA instruction usage
+
+---
+
+## File Locations Reference
+
+### Main Device Operation Headers
+```
+library/include/ck/library/tensor_operation_instance/gpu/
+├── grouped_conv_fwd/
+│   ├── device_grouped_conv_fwd_xdl_instance.hpp          ← PRIMARY FILE
+│   ├── device_grouped_conv_fwd_dl_instance.hpp
+│   ├── device_grouped_conv_fwd_wmma_instance.hpp
+│   ├── device_grouped_conv_fwd_xdl_comp_instance.hpp
+│   ├── device_grouped_conv_fwd_xdl_mem_instance.hpp
+│   ├── device_grouped_conv_fwd_xdl_large_tensor_instance.hpp
+│   └── ... (other variants)
+└── convolution_forward.hpp                               ← INTERFACE DEFINITIONS
+```
+
+### Source Instantiations
+```
+library/src/tensor_operation_instance/gpu/
+├── conv2d_fwd/                    ← Non-grouped 2D convolutions
+├── grouped_conv1d_fwd/            ← 1D grouped convolutions
+├── grouped_conv2d_fwd/            ← 2D grouped convolutions
+│   ├── dl/                        ← Direct Load variants
+│   ├── wmma/                      ← WMMA variants
+│   └── xdl/                       ← XDL variants
+│       ├── comp/                  ← Compute-optimized
+│       ├── mem/                   ← Memory-optimized
+│       ├── large_tensor/          ← Large tensor optimized
+│       └── merged_groups/         ← Merged groups optimized
+└── grouped_conv3d_fwd/            ← 3D grouped convolutions
+```
+
+---
+
+## Additional Resources
+
+For detailed analysis and full template instantiations, refer to the generated reports:
+- `forward_convolution_device_ops_report.txt` - Quick summary
+- `forward_conv_device_ops_detailed_report.txt` - Full details with examples
+- `forward_convolution_device_ops_data.json` - Machine-readable format
+
+The analysis scripts used to generate this information:
+- `extract_conv_fwd_device_ops.py` - Initial device operation extraction
+- `extract_detailed_instantiations.py` - Detailed instantiation analysis

extract_all_instantiations_complete.py

@@ -0,0 +1,296 @@
+#!/usr/bin/env python3
+"""
+Extract ALL template instantiations for forward convolution device operations.
+Generates both human-readable .txt and programmatic JSON outputs.
+"""
+
+import os
+import re
+from pathlib import Path
+from collections import defaultdict
+import json
+
+def find_files(base_paths, extensions):
+    """Find all files with given extensions in base paths."""
+    files = []
+    for base_path in base_paths:
+        path_obj = Path(base_path)
+        if path_obj.is_dir():
+            for ext in extensions:
+                files.extend(path_obj.rglob(f'*{ext}'))
+    return files
+
+def extract_complete_instantiation(lines, start_idx, device_op_name):
+    """Extract a complete device operation instantiation with balanced brackets."""
+    if start_idx >= len(lines):
+        return None, start_idx
+    
+    line = lines[start_idx]
+    
+    # Check if this line contains the device op
+    if device_op_name + '<' not in line:
+        return None, start_idx
+    
+    # Start collecting the full instantiation
+    full_lines = [line]
+    bracket_count = line.count('<') - line.count('>')
+    
+    # Continue to next lines until brackets are balanced
+    idx = start_idx + 1
+    while bracket_count > 0 and idx < len(lines):
+        full_lines.append(lines[idx])
+        bracket_count += lines[idx].count('<') - lines[idx].count('>')
+        idx += 1
+    
+    # Join and clean up
+    full_text = '\n'.join(full_lines)
+    
+    return {
+        'line_start': start_idx + 1,
+        'line_end': idx,
+        'full_text': full_text.strip(),
+        'compressed_text': ' '.join(full_text.split())  # Single line version
+    }, idx
+
+def extract_all_instantiations(content, device_op_name):
+    """Extract ALL device operation instantiations from content."""
+    instantiations = []
+    lines = content.split('\n')
+    
+    idx = 0
+    while idx < len(lines):
+        inst, next_idx = extract_complete_instantiation(lines, idx, device_op_name)
+        if inst:
+            instantiations.append(inst)
+            idx = next_idx
+        else:
+            idx += 1
+    
+    return instantiations
+
+def parse_template_parameters(instantiation_text):
+    """Parse template parameters from instantiation text (simplified)."""
+    # This is a simplified parser - actual parsing would need full C++ template parser
+    # For now, just extract key information
+    params = {}
+    
+    # Extract data types
+    type_patterns = {
+        'BF16': r'\bBF16\b',
+        'F16': r'\bF16\b',
+        'F32': r'\bF32\b',
+        'TF32': r'\bTF32\b',
+        'INT8': r'\bint8_t\b',
+        'F8': r'\bF8\b',
+        'BF8': r'\bBF8\b',
+    }
+    
+    for type_name, pattern in type_patterns.items():
+        if re.search(pattern, instantiation_text):
+            params[f'uses_{type_name}'] = True
+    
+    # Extract block sizes if visible
+    block_size_match = re.search(r',\s*(\d{2,3}),\s*(\d{2,3}),\s*(\d{2,3}),\s*(\d{1,3}),', instantiation_text)
+    if block_size_match:
+        params['block_size'] = block_size_match.group(1)
+        params['m_per_block'] = block_size_match.group(2)
+        params['n_per_block'] = block_size_match.group(3)
+        params['k_per_block'] = block_size_match.group(4)
+    
+    return params
+
+def main():
+    # Device operations to extract
+    target_device_ops = [
+        'DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle',
+        'DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3',
+        'DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K',
+        'DeviceConv2dFwdXdl_C_Shuffle_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K',
+        'DeviceGroupedConvFwdMultipleD_Wmma_CShuffle',
+        'DeviceGroupedConvFwdDlMultipleD_NHWC_KYXC_NHWK',
+        'DeviceGroupedConvFwdMultipleD_Xdl_CShuffle_Large_Tensor',
+        'DeviceConvFwd',
+    ]
+    
+    # Define paths
+    header_paths = [
+        'library/include/ck/library/tensor_operation_instance/gpu/grouped_conv_fwd',
+        'library/include/ck/library/tensor_operation_instance/gpu',
+    ]
+    
+    source_paths = [
+        'library/src/tensor_operation_instance/gpu/grouped_conv1d_fwd',
+        'library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd',
+        'library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd',
+        'library/src/tensor_operation_instance/gpu/conv2d_fwd',
+    ]
+    
+    # Find all files
+    header_files = find_files(header_paths, ['.hpp', '.h', '.inc'])
+    source_files = find_files(source_paths, ['.cpp', '.hpp', '.inc', '.in'])
+    all_files = list(set(header_files + source_files))
+    
+    print(f"Analyzing {len(all_files)} files for ALL instantiations...\n")
+    
+    # Collect data
+    all_results = {}
+    
+    for device_op in target_device_ops:
+        print(f"Extracting all instantiations for {device_op}...")
+        
+        device_op_data = {
+            'device_operation_name': device_op,
+            'files': {},
+            'total_instantiations': 0
+        }
+        
+        for file_path in all_files:
+            try:
+                with open(file_path, 'r', encoding='utf-8', errors='ignore') as f:
+                    content = f.read()
+                
+                if device_op in content:
+                    instantiations = extract_all_instantiations(content, device_op)
+                    
+                    if instantiations:
+                        rel_path = str(file_path).replace(os.getcwd() + '/', '')
+                        
+                        # Process each instantiation
+                        processed_insts = []
+                        for inst in instantiations:
+                            processed_inst = {
+                                'line_start': inst['line_start'],
+                                'line_end': inst['line_end'],
+                                'full_text': inst['full_text'],
+                                'compressed_text': inst['compressed_text'],
+                                'template_params': parse_template_parameters(inst['full_text'])
+                            }
+                            processed_insts.append(processed_inst)
+                        
+                        device_op_data['files'][rel_path] = processed_insts
+                        device_op_data['total_instantiations'] += len(instantiations)
+            
+            except Exception as e:
+                print(f"  Error processing {file_path}: {e}")
+        
+        all_results[device_op] = device_op_data
+        print(f"  Found {device_op_data['total_instantiations']} instantiations in {len(device_op_data['files'])} files")
+    
+    # Generate comprehensive TXT report
+    print("\nGenerating comprehensive TXT report...")
+    txt_report = []
+    txt_report.append("=" * 120)
+    txt_report.append("COMPLETE FORWARD CONVOLUTION DEVICE OPERATIONS AND ALL TEMPLATE INSTANTIATIONS")
+    txt_report.append("=" * 120)
+    txt_report.append("")
+    txt_report.append(f"Namespace: ck::tensor_operation::device")
+    txt_report.append(f"Total Device Operation Types: {len(target_device_ops)}")
+    total_insts = sum(data['total_instantiations'] for data in all_results.values())
+    txt_report.append(f"Total Template Instantiations: {total_insts}")
+    txt_report.append("")
+    
+    # Summary table
+    txt_report.append("SUMMARY OF DEVICE OPERATIONS")
+    txt_report.append("-" * 120)
+    for device_op in target_device_ops:
+        data = all_results[device_op]
+        txt_report.append(f"{data['total_instantiations']:4d} instantiations | {len(data['files']):3d} files | {device_op}")
+    txt_report.append("")
+    txt_report.append("=" * 120)
+    txt_report.append("")
+    
+    # Detailed section for each device op - ALL instantiations
+    for device_op in target_device_ops:
+        data = all_results[device_op]
+        
+        txt_report.append("")
+        txt_report.append("=" * 120)
+        txt_report.append(f"DEVICE OPERATION: {device_op}")
+        txt_report.append("=" * 120)
+        txt_report.append("")
+        txt_report.append(f"Total Instantiations: {data['total_instantiations']}")
+        txt_report.append(f"Number of Files: {len(data['files'])}")
+        txt_report.append("")
+        
+        # List ALL instantiations in each file
+        for file_path, instantiations in sorted(data['files'].items()):
+            txt_report.append("-" * 120)
+            txt_report.append(f"FILE: {file_path}")
+            txt_report.append(f"Instantiations: {len(instantiations)}")
+            txt_report.append("")
+            
+            for idx, inst in enumerate(instantiations, 1):
+                txt_report.append(f"[{idx}] Lines {inst['line_start']}-{inst['line_end']}:")
+                txt_report.append("")
+                # Include full text with proper indentation
+                for line in inst['full_text'].split('\n'):
+                    txt_report.append(f"    {line}")
+                txt_report.append("")
+            
+            txt_report.append("")
+        
+        txt_report.append("")
+    
+    # Save TXT report
+    txt_output = 'forward_conv_all_instantiations.txt'
+    with open(txt_output, 'w') as f:
+        f.write('\n'.join(txt_report))
+    print(f"✓ TXT report saved: {txt_output}")
+    
+    # Generate JSON for programmatic use
+    print("Generating JSON for programmatic use...")
+    json_data = {
+        'metadata': {
+            'description': 'Forward convolution device operations and template instantiations',
+            'namespace': 'ck::tensor_operation::device',
+            'total_device_operations': len(target_device_ops),
+            'total_instantiations': total_insts
+        },
+        'device_operations': {}
+    }
+    
+    for device_op in target_device_ops:
+        data = all_results[device_op]
+        
+        json_device_op = {
+            'name': device_op,
+            'total_instantiations': data['total_instantiations'],
+            'total_files': len(data['files']),
+            'instantiations_by_file': {}
+        }
+        
+        for file_path, instantiations in data['files'].items():
+            file_insts = []
+            for inst in instantiations:
+                file_insts.append({
+                    'line_start': inst['line_start'],
+                    'line_end': inst['line_end'],
+                    'instantiation_text': inst['full_text'],
+                    'instantiation_compressed': inst['compressed_text'],
+                    'parsed_parameters': inst['template_params']
+                })
+            
+            json_device_op['instantiations_by_file'][file_path] = {
+                'count': len(file_insts),
+                'instantiations': file_insts
+            }
+        
+        json_data['device_operations'][device_op] = json_device_op
+    
+    # Save JSON
+    json_output = 'forward_conv_all_instantiations.json'
+    with open(json_output, 'w') as f:
+        json.dump(json_data, f, indent=2)
+    print(f"✓ JSON file saved: {json_output}")
+    
+    # Summary
+    print(f"\n{'='*60}")
+    print(f"GENERATION COMPLETE")
+    print(f"{'='*60}")
+    print(f"TXT File:  {txt_output}")
+    print(f"JSON File: {json_output}")
+    print(f"Total instantiations extracted: {total_insts}")
+    print(f"{'='*60}")
+
+if __name__ == '__main__':
+    main()

extract_backward_conv_ops.py

@@ -0,0 +1,484 @@
+#!/usr/bin/env python3
+"""
+Extract ALL backward convolution device operations and template instantiations.
+Handles both backward weight and backward data convolutions.
+Generates human-readable .txt and programmatic JSON outputs.
+"""
+
+import os
+import re
+from pathlib import Path
+from collections import defaultdict
+import json
+
+def find_files(base_paths, extensions):
+    """Find all files with given extensions in base paths."""
+    files = []
+    for base_path in base_paths:
+        path_obj = Path(base_path)
+        if path_obj.is_dir():
+            for ext in extensions:
+                files.extend(path_obj.rglob(f'*{ext}'))
+    return files
+
+def extract_complete_instantiation(lines, start_idx, device_op_name):
+    """Extract a complete device operation instantiation with balanced brackets."""
+    if start_idx >= len(lines):
+        return None, start_idx
+    
+    line = lines[start_idx]
+    
+    # Check if this line contains the device op
+    if device_op_name + '<' not in line:
+        return None, start_idx
+    
+    # Start collecting the full instantiation
+    full_lines = [line]
+    bracket_count = line.count('<') - line.count('>')
+    
+    # Continue to next lines until brackets are balanced
+    idx = start_idx + 1
+    while bracket_count > 0 and idx < len(lines):
+        full_lines.append(lines[idx])
+        bracket_count += lines[idx].count('<') - lines[idx].count('>')
+        idx += 1
+    
+    # Join and clean up
+    full_text = '\n'.join(full_lines)
+    
+    return {
+        'line_start': start_idx + 1,
+        'line_end': idx,
+        'full_text': full_text.strip(),
+        'compressed_text': ' '.join(full_text.split())
+    }, idx
+
+def extract_all_instantiations(content, device_op_name):
+    """Extract ALL device operation instantiations from content."""
+    instantiations = []
+    lines = content.split('\n')
+    
+    idx = 0
+    while idx < len(lines):
+        inst, next_idx = extract_complete_instantiation(lines, idx, device_op_name)
+        if inst:
+            instantiations.append(inst)
+            idx = next_idx
+        else:
+            idx += 1
+    
+    return instantiations
+
+def parse_template_parameters(instantiation_text):
+    """Parse template parameters from instantiation text."""
+    params = {}
+    
+    # Extract data types
+    type_patterns = {
+        'BF16': r'\bBF16\b',
+        'F16': r'\bF16\b',
+        'F32': r'\bF32\b',
+        'TF32': r'\bTF32\b',
+        'INT8': r'\bint8_t\b',
+        'F8': r'\bF8\b',
+        'BF8': r'\bBF8\b',
+    }
+    
+    for type_name, pattern in type_patterns.items():
+        if re.search(pattern, instantiation_text):
+            params[f'uses_{type_name}'] = True
+    
+    # Extract block sizes if visible
+    block_size_match = re.search(r',\s*(\d{2,3}),\s*(\d{2,3}),\s*(\d{2,3}),\s*(\d{1,3}),', instantiation_text)
+    if block_size_match:
+        params['block_size'] = block_size_match.group(1)
+        params['m_per_block'] = block_size_match.group(2)
+        params['n_per_block'] = block_size_match.group(3)
+        params['k_per_block'] = block_size_match.group(4)
+    
+    return params
+
+def find_device_operations(all_files):
+    """Find all device operations matching backward convolution patterns."""
+    device_ops = set()
+    
+    # Patterns for backward operations
+    patterns = [
+        r'(Device\w*Conv\w*Bwd\w*Weight\w*)<',
+        r'(Device\w*Conv\w*Bwd\w*Data\w*)<',
+        r'(Device\w*ConvBwd\w*)<',
+    ]
+    
+    for file_path in all_files:
+        try:
+            with open(file_path, 'r', encoding='utf-8', errors='ignore') as f:
+                content = f.read()
+            
+            for pattern in patterns:
+                matches = re.finditer(pattern, content)
+                for match in matches:
+                    device_ops.add(match.group(1))
+        except:
+            pass
+    
+    return sorted(list(device_ops))
+
+def main():
+    print("="*80)
+    print("EXTRACTING BACKWARD CONVOLUTION DEVICE OPERATIONS")
+    print("="*80)
+    print()
+    
+    # Define paths
+    header_paths = [
+        'library/include/ck/library/tensor_operation_instance/gpu/grouped_conv_bwd_weight',
+        'library/include/ck/library/tensor_operation_instance/gpu/grouped_conv_bwd_data',
+        'library/include/ck/library/tensor_operation_instance/gpu',
+    ]
+    
+    source_paths = [
+        'library/src/tensor_operation_instance/gpu/grouped_conv1d_bwd_weight',
+        'library/src/tensor_operation_instance/gpu/grouped_conv2d_bwd_weight',
+        'library/src/tensor_operation_instance/gpu/grouped_conv3d_bwd_weight',
+        'library/src/tensor_operation_instance/gpu/grouped_convnd_bwd_weight',
+        'library/src/tensor_operation_instance/gpu/conv1d_bwd_data',
+        'library/src/tensor_operation_instance/gpu/conv2d_bwd_data',
+        'library/src/tensor_operation_instance/gpu/conv3d_bwd_data',
+        'library/src/tensor_operation_instance/gpu/grouped_conv1d_bwd_data',
+        'library/src/tensor_operation_instance/gpu/grouped_conv2d_bwd_data',
+        'library/src/tensor_operation_instance/gpu/grouped_conv3d_bwd_data',
+    ]
+    
+    # Find all files
+    header_files = find_files(header_paths, ['.hpp', '.h', '.inc'])
+    source_files = find_files(source_paths, ['.cpp', '.hpp', '.inc', '.in'])
+    all_files = list(set(header_files + source_files))
+    
+    print(f"Found {len(all_files)} files to analyze")
+    
+    # Discover all backward device operations
+    print("Discovering device operations...")
+    target_device_ops = find_device_operations(all_files)
+    print(f"Found {len(target_device_ops)} device operation types")
+    for op in target_device_ops:
+        print(f"  - {op}")
+    print()
+    
+    # Collect data
+    all_results = {}
+    
+    for device_op in target_device_ops:
+        print(f"Extracting all instantiations for {device_op}...")
+        
+        device_op_data = {
+            'device_operation_name': device_op,
+            'files': {},
+            'total_instantiations': 0
+        }
+        
+        for file_path in all_files:
+            try:
+                with open(file_path, 'r', encoding='utf-8', errors='ignore') as f:
+                    content = f.read()
+                
+                if device_op in content:
+                    instantiations = extract_all_instantiations(content, device_op)
+                    
+                    if instantiations:
+                        rel_path = str(file_path).replace(os.getcwd() + '/', '')
+                        
+                        processed_insts = []
+                        for inst in instantiations:
+                            processed_inst = {
+                                'line_start': inst['line_start'],
+                                'line_end': inst['line_end'],
+                                'full_text': inst['full_text'],
+                                'compressed_text': inst['compressed_text'],
+                                'template_params': parse_template_parameters(inst['full_text'])
+                            }
+                            processed_insts.append(processed_inst)
+                        
+                        device_op_data['files'][rel_path] = processed_insts
+                        device_op_data['total_instantiations'] += len(instantiations)
+            
+            except Exception as e:
+                pass
+        
+        all_results[device_op] = device_op_data
+        print(f"  Found {device_op_data['total_instantiations']} instantiations in {len(device_op_data['files'])} files")
+    
+    # Separate by type
+    bwd_weight_ops = {k: v for k, v in all_results.items() if 'Weight' in k or 'Wgt' in k}
+    bwd_data_ops = {k: v for k, v in all_results.items() if 'Data' in k}
+    
+    print(f"\nBackward Weight Operations: {len(bwd_weight_ops)}")
+    print(f"Backward Data Operations: {len(bwd_data_ops)}")
+    
+    # Generate TXT report
+    print("\nGenerating comprehensive TXT report...")
+    txt_report = generate_txt_report(all_results, bwd_weight_ops, bwd_data_ops)
+    
+    txt_output = 'backward_conv_all_instantiations.txt'
+    with open(txt_output, 'w') as f:
+        f.write(txt_report)
+    print(f"✓ TXT report saved: {txt_output}")
+    
+    # Generate JSON
+    print("Generating JSON for programmatic use...")
+    json_data = generate_json_data(all_results, bwd_weight_ops, bwd_data_ops)
+    
+    json_output = 'backward_conv_all_instantiations.json'
+    with open(json_output, 'w') as f:
+        json.dump(json_data, f, indent=2)
+    print(f"✓ JSON file saved: {json_output}")
+    
+    # Generate markdown summary
+    print("Generating markdown summary...")
+    md_content = generate_markdown_summary(all_results, bwd_weight_ops, bwd_data_ops)
+    
+    md_output = 'BACKWARD_CONVOLUTION_DEVICE_OPS_SUMMARY.md'
+    with open(md_output, 'w') as f:
+        f.write(md_content)
+    print(f"✓ Markdown summary saved: {md_output}")
+    
+    # Final summary
+    total_insts = sum(data['total_instantiations'] for data in all_results.values())
+    print(f"\n{'='*80}")
+    print(f"GENERATION COMPLETE")
+    print(f"{'='*80}")
+    print(f"Device Operations Found: {len(all_results)}")
+    print(f"  - Backward Weight: {len(bwd_weight_ops)}")
+    print(f"  - Backward Data: {len(bwd_data_ops)}")
+    print(f"Total Instantiations: {total_insts}")
+    print(f"\nFiles generated:")
+    print(f"  - {txt_output}")
+    print(f"  - {json_output}")
+    print(f"  - {md_output}")
+    print(f"{'='*80}")
+
+def generate_txt_report(all_results, bwd_weight_ops, bwd_data_ops):
+    """Generate comprehensive TXT report."""
+    report = []
+    report.append("=" * 120)
+    report.append("COMPLETE BACKWARD CONVOLUTION DEVICE OPERATIONS AND ALL TEMPLATE INSTANTIATIONS")
+    report.append("=" * 120)
+    report.append("")
+    report.append(f"Namespace: ck::tensor_operation::device")
+    report.append(f"Total Device Operation Types: {len(all_results)}")
+    total_insts = sum(data['total_instantiations'] for data in all_results.values())
+    report.append(f"Total Template Instantiations: {total_insts}")
+    report.append("")
+    
+    # Summary
+    report.append("SUMMARY OF DEVICE OPERATIONS")
+    report.append("-" * 120)
+    report.append("")
+    report.append("BACKWARD WEIGHT OPERATIONS:")
+    for device_op, data in sorted(bwd_weight_ops.items(), key=lambda x: x[1]['total_instantiations'], reverse=True):
+        report.append(f" {data['total_instantiations']:4d} instantiations | {len(data['files']):3d} files | {device_op}")
+    report.append("")
+    report.append("BACKWARD DATA OPERATIONS:")
+    for device_op, data in sorted(bwd_data_ops.items(), key=lambda x: x[1]['total_instantiations'], reverse=True):
+        report.append(f" {data['total_instantiations']:4d} instantiations | {len(data['files']):3d} files | {device_op}")
+    report.append("")
+    report.append("=" * 120)
+    report.append("")
+    
+    # Detailed sections
+    for category_name, ops_dict in [("BACKWARD WEIGHT", bwd_weight_ops), ("BACKWARD DATA", bwd_data_ops)]:
+        report.append("")
+        report.append("=" * 120)
+        report.append(f"{category_name} OPERATIONS")
+        report.append("=" * 120)
+        report.append("")
+        
+        for device_op in sorted(ops_dict.keys()):
+            data = all_results[device_op]
+            
+            report.append("=" * 120)
+            report.append(f"DEVICE OPERATION: {device_op}")
+            report.append("=" * 120)
+            report.append("")
+            report.append(f"Total Instantiations: {data['total_instantiations']}")
+            report.append(f"Number of Files: {len(data['files'])}")
+            report.append("")
+            
+            for file_path, instantiations in sorted(data['files'].items()):
+                report.append("-" * 120)
+                report.append(f"FILE: {file_path}")
+                report.append(f"Instantiations: {len(instantiations)}")
+                report.append("")
+                
+                for idx, inst in enumerate(instantiations, 1):
+                    report.append(f"[{idx}] Lines {inst['line_start']}-{inst['line_end']}:")
+                    report.append("")
+                    for line in inst['full_text'].split('\n'):
+                        report.append(f"    {line}")
+                    report.append("")
+                
+                report.append("")
+            
+            report.append("")
+    
+    return '\n'.join(report)
+
+def generate_json_data(all_results, bwd_weight_ops, bwd_data_ops):
+    """Generate JSON data for programmatic use."""
+    total_insts = sum(data['total_instantiations'] for data in all_results.values())
+    
+    json_data = {
+        'metadata': {
+            'description': 'Backward convolution device operations and template instantiations',
+            'namespace': 'ck::tensor_operation::device',
+            'total_device_operations': len(all_results),
+            'backward_weight_operations': len(bwd_weight_ops),
+            'backward_data_operations': len(bwd_data_ops),
+            'total_instantiations': total_insts
+        },
+        'backward_weight_operations': {},
+        'backward_data_operations': {}
+    }
+    
+    # Process backward weight operations
+    for device_op, data in bwd_weight_ops.items():
+        json_device_op = {
+            'name': device_op,
+            'total_instantiations': data['total_instantiations'],
+            'total_files': len(data['files']),
+            'instantiations_by_file': {}
+        }
+        
+        for file_path, instantiations in data['files'].items():
+            file_insts = []
+            for inst in instantiations:
+                file_insts.append({
+                    'line_start': inst['line_start'],
+                    'line_end': inst['line_end'],
+                    'instantiation_text': inst['full_text'],
+                    'instantiation_compressed': inst['compressed_text'],
+                    'parsed_parameters': inst['template_params']
+                })
+            
+            json_device_op['instantiations_by_file'][file_path] = {
+                'count': len(file_insts),
+                'instantiations': file_insts
+            }
+        
+        json_data['backward_weight_operations'][device_op] = json_device_op
+    
+    # Process backward data operations
+    for device_op, data in bwd_data_ops.items():
+        json_device_op = {
+            'name': device_op,
+            'total_instantiations': data['total_instantiations'],
+            'total_files': len(data['files']),
+            'instantiations_by_file': {}
+        }
+        
+        for file_path, instantiations in data['files'].items():
+            file_insts = []
+            for inst in instantiations:
+                file_insts.append({
+                    'line_start': inst['line_start'],
+                    'line_end': inst['line_end'],
+                    'instantiation_text': inst['full_text'],
+                    'instantiation_compressed': inst['compressed_text'],
+                    'parsed_parameters': inst['template_params']
+                })
+            
+            json_device_op['instantiations_by_file'][file_path] = {
+                'count': len(file_insts),
+                'instantiations': file_insts
+            }
+        
+        json_data['backward_data_operations'][device_op] = json_device_op
+    
+    return json_data
+
+def generate_markdown_summary(all_results, bwd_weight_ops, bwd_data_ops):
+    """Generate markdown summary document."""
+    total_insts = sum(data['total_instantiations'] for data in all_results.values())
+    bwd_weight_insts = sum(data['total_instantiations'] for data in bwd_weight_ops.values())
+    bwd_data_insts = sum(data['total_instantiations'] for data in bwd_data_ops.values())
+    
+    md = []
+    md.append("# Backward Convolution Device Operations - Comprehensive Summary")
+    md.append("")
+    md.append("**Generated:** October 29, 2025  ")
+    md.append("**Namespace:** `ck::tensor_operation::device`  ")
+    md.append("**Location:** `/library/include` and `/library/src` directories")
+    md.append("")
+    md.append("---")
+    md.append("")
+    md.append("## Overview")
+    md.append("")
+    md.append("This document provides a comprehensive list of all device operations used for backward convolutions (both weight gradients and data gradients) in the Composable Kernel library.")
+    md.append("")
+    md.append("### Total Statistics")
+    md.append(f"- **Unique Device Operation Types:** {len(all_results)}")
+    md.append(f"- **Backward Weight Operations:** {len(bwd_weight_ops)} types ({bwd_weight_insts} instantiations)")
+    md.append(f"- **Backward Data Operations:** {len(bwd_data_ops)} types ({bwd_data_insts} instantiations)")
+    md.append(f"- **Total Template Instantiations:** {total_insts}")
+    md.append("")
+    md.append("---")
+    md.append("")
+    
+    # Backward Weight Table
+    md.append("## Backward Weight Device Operations")
+    md.append("")
+    md.append("| # | Device Operation Name | Instantiations | Files | Primary Location |")
+    md.append("|---|----------------------|----------------|-------|------------------|")
+    
+    for idx, (device_op, data) in enumerate(sorted(bwd_weight_ops.items(), key=lambda x: x[1]['total_instantiations'], reverse=True), 1):
+        primary_file = sorted(data['files'].keys())[0] if data['files'] else "N/A"
+        primary_file_short = primary_file.split('/')[-1] if '/' in primary_file else primary_file
+        md.append(f"| {idx} | `{device_op}` | {data['total_instantiations']} | {len(data['files'])} | `{primary_file_short}` |")
+    
+    md.append(f"| **Total** | **Backward Weight Operations** | **{bwd_weight_insts}** | **{sum(len(d['files']) for d in bwd_weight_ops.values())}** | |")
+    md.append("")
+    md.append("---")
+    md.append("")
+    
+    # Backward Data Table
+    md.append("## Backward Data Device Operations")
+    md.append("")
+    md.append("| # | Device Operation Name | Instantiations | Files | Primary Location |")
+    md.append("|---|----------------------|----------------|-------|------------------|")
+    
+    for idx, (device_op, data) in enumerate(sorted(bwd_data_ops.items(), key=lambda x: x[1]['total_instantiations'], reverse=True), 1):
+        primary_file = sorted(data['files'].keys())[0] if data['files'] else "N/A"
+        primary_file_short = primary_file.split('/')[-1] if '/' in primary_file else primary_file
+        md.append(f"| {idx} | `{device_op}` | {data['total_instantiations']} | {len(data['files'])} | `{primary_file_short}` |")
+    
+    md.append(f"| **Total** | **Backward Data Operations** | **{bwd_data_insts}** | **{sum(len(d['files']) for d in bwd_data_ops.values())}** | |")
+    md.append("")
+    md.append(f"### Grand Total: {total_insts} Template Instantiations across {len(all_results)} Device Operation Types")
+    md.append("")
+    md.append("---")
+    md.append("")
+    
+    # Output files section
+    md.append("## Output Files")
+    md.append("")
+    md.append("### Complete Instantiation Files")
+    md.append("")
+    md.append("1. **`backward_conv_all_instantiations.txt`**")
+    md.append("   - COMPLETE listing of ALL template instantiations")
+    md.append("   - Human-readable format with line numbers")
+    md.append("   - Organized by operation type, then by file")
+    md.append("")
+    md.append("2. **`backward_conv_all_instantiations.json`**")
+    md.append("   - Structured JSON for programmatic instantiation generation")
+    md.append("   - Separated into `backward_weight_operations` and `backward_data_operations`")
+    md.append("   - Each instantiation includes full text and parsed parameters")
+    md.append("   - Ready for automated code generation")
+    md.append("")
+    md.append("3. **`BACKWARD_CONVOLUTION_DEVICE_OPS_SUMMARY.md`** (This file)")
+    md.append("   - Executive summary with tables")
+    md.append("   - Quick reference for all backward operations")
+    md.append("")
+    
+    return '\n'.join(md)
+
+if __name__ == '__main__':
+    main()

extract_conv_fwd_device_ops.py

@@ -0,0 +1,195 @@
+#!/usr/bin/env python3
+"""
+Extract all forward convolution device operations and their template instantiations
+from CK library headers and source files.
+"""
+
+import os
+import re
+from pathlib import Path
+from collections import defaultdict
+import json
+
+# Device operation patterns to search for
+DEVICE_OP_PATTERNS = [
+    r'(Device\w*Conv\w*Fwd\w*)<',  # Generic pattern for device ops
+]
+
+def find_files(base_paths, extensions):
+    """Find all files with given extensions in base paths."""
+    files = []
+    for base_path in base_paths:
+        path_obj = Path(base_path)
+        if path_obj.is_dir():
+            for ext in extensions:
+                files.extend(path_obj.rglob(f'*{ext}'))
+    return files
+
+def extract_device_operations(file_path):
+    """Extract device operations from a file."""
+    device_ops = []
+    
+    try:
+        with open(file_path, 'r', encoding='utf-8', errors='ignore') as f:
+            content = f.read()
+            
+        # Search for device operation class names
+        for pattern in DEVICE_OP_PATTERNS:
+            matches = re.finditer(pattern, content)
+            for match in matches:
+                device_op_name = match.group(1)
+                # Store the device operation with file location
+                device_ops.append({
+                    'name': device_op_name,
+                    'file': str(file_path),
+                    'line_context': None  # We'll extract context later if needed
+                })
+    except Exception as e:
+        print(f"Error processing {file_path}: {e}")
+    
+    return device_ops
+
+def extract_template_instantiation(content, device_op_name):
+    """Extract a single template instantiation for analysis."""
+    # Look for the full template instantiation
+    pattern = rf'{re.escape(device_op_name)}<[^>]*(?:<[^>]*>)*[^>]*>'
+    
+    # This is complex due to nested templates, so let's use a simpler approach
+    # Just find lines with the device op name
+    lines = content.split('\n')
+    instantiations = []
+    
+    for i, line in enumerate(lines):
+        if device_op_name in line and '<' in line:
+            # Try to capture the full instantiation (might span multiple lines)
+            instantiation_text = line.strip()
+            
+            # Count template brackets to see if we have a complete instantiation
+            open_brackets = instantiation_text.count('<')
+            close_brackets = instantiation_text.count('>')
+            
+            # If brackets are balanced, we have a complete line
+            if open_brackets > 0:
+                instantiations.append({
+                    'line_number': i + 1,
+                    'text': instantiation_text[:200]  # Truncate for readability
+                })
+    
+    return instantiations
+
+def main():
+    # Define paths to search
+    header_paths = [
+        'library/include/ck/library/tensor_operation_instance/gpu/grouped_conv_fwd',
+        'library/include/ck/library/tensor_operation_instance/gpu',
+    ]
+    
+    source_paths = [
+        'library/src/tensor_operation_instance/gpu/grouped_conv1d_fwd',
+        'library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd',
+        'library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd',
+        'library/src/tensor_operation_instance/gpu/conv2d_fwd',
+    ]
+    
+    all_paths = header_paths + source_paths
+    
+    # Find all relevant files
+    header_files = find_files(header_paths, ['.hpp', '.h'])
+    source_files = find_files(source_paths, ['.cpp', '.hpp', '.inc', '.in'])
+    
+    all_files = list(set(header_files + source_files))
+    
+    print(f"Found {len(all_files)} files to analyze")
+    
+    # Collect all device operations
+    all_device_ops = []
+    for file_path in all_files:
+        ops = extract_device_operations(file_path)
+        all_device_ops.extend(ops)
+    
+    # Group by device operation name
+    ops_by_name = defaultdict(list)
+    for op in all_device_ops:
+        ops_by_name[op['name']].append(op)
+    
+    # Count unique device operations
+    unique_ops = sorted(ops_by_name.keys())
+    
+    print(f"\nFound {len(unique_ops)} unique device operation types:")
+    for op_name in unique_ops:
+        count = len(ops_by_name[op_name])
+        print(f"  {op_name}: {count} occurrences")
+    
+    # Now extract detailed instantiations for each device op
+    device_op_details = {}
+    
+    for device_op_name in unique_ops:
+        instantiation_files = set()
+        total_instantiations = 0
+        
+        # Re-scan files that contain this device op
+        for file_path in all_files:
+            try:
+                with open(file_path, 'r', encoding='utf-8', errors='ignore') as f:
+                    content = f.read()
+                
+                if device_op_name in content:
+                    instantiations = extract_template_instantiation(content, device_op_name)
+                    if instantiations:
+                        instantiation_files.add(str(file_path))
+                        total_instantiations += len(instantiations)
+            except Exception as e:
+                pass
+        
+        device_op_details[device_op_name] = {
+            'files': sorted(list(instantiation_files)),
+            'total_instantiations': total_instantiations
+        }
+    
+    # Generate report
+    report = []
+    report.append("=" * 80)
+    report.append("FORWARD CONVOLUTION DEVICE OPERATIONS REPORT")
+    report.append("=" * 80)
+    report.append("")
+    report.append(f"Total Unique Device Operations: {len(unique_ops)}")
+    report.append("")
+    
+    for device_op_name in unique_ops:
+        details = device_op_details[device_op_name]
+        report.append("-" * 80)
+        report.append(f"Device Operation: {device_op_name}")
+        report.append(f"Total Template Instantiations: {details['total_instantiations']}")
+        report.append(f"Number of Files: {len(details['files'])}")
+        report.append("")
+        report.append("Files containing this device operation:")
+        for file_path in details['files']:
+            # Make path relative to workspace
+            rel_path = file_path.replace(os.getcwd() + '/', '')
+            report.append(f"  - {rel_path}")
+        report.append("")
+    
+    # Save report
+    report_text = '\n'.join(report)
+    
+    with open('forward_convolution_device_ops_report.txt', 'w') as f:
+        f.write(report_text)
+    
+    print(f"\nReport saved to: forward_convolution_device_ops_report.txt")
+    
+    # Also save as JSON for programmatic access
+    json_data = {
+        'summary': {
+            'total_device_operations': len(unique_ops),
+            'device_operation_names': unique_ops
+        },
+        'details': device_op_details
+    }
+    
+    with open('forward_convolution_device_ops_data.json', 'w') as f:
+        json.dump(json_data, f, indent=2)
+    
+    print(f"JSON data saved to: forward_convolution_device_ops_data.json")
+
+if __name__ == '__main__':
+    main()

extract_detailed_instantiations.py

@@ -0,0 +1,201 @@
+#!/usr/bin/env python3
+"""
+Extract detailed template instantiations for forward convolution device operations.
+This script parses the actual template parameters for each device operation instance.
+"""
+
+import os
+import re
+from pathlib import Path
+from collections import defaultdict
+
+def find_files(base_paths, extensions):
+    """Find all files with given extensions in base paths."""
+    files = []
+    for base_path in base_paths:
+        path_obj = Path(base_path)
+        if path_obj.is_dir():
+            for ext in extensions:
+                files.extend(path_obj.rglob(f'*{ext}'))
+    return files
+
+def extract_device_op_instantiations(content, device_op_name):
+    """Extract complete device operation instantiations."""
+    instantiations = []
+    lines = content.split('\n')
+    
+    i = 0
+    while i < len(lines):
+        line = lines[i]
+        
+        # Check if this line contains the device op
+        if device_op_name + '<' in line:
+            # Start collecting the full instantiation
+            full_instantiation = line
+            bracket_count = line.count('<') - line.count('>')
+            
+            # Continue to next lines if brackets aren't balanced
+            j = i + 1
+            while bracket_count > 0 and j < len(lines):
+                full_instantiation += '\n' + lines[j]
+                bracket_count += lines[j].count('<') - lines[j].count('>')
+                j += 1
+            
+            # Clean up and store
+            instantiation_clean = full_instantiation.strip()
+            if instantiation_clean:
+                instantiations.append({
+                    'line_start': i + 1,
+                    'line_end': j,
+                    'text': instantiation_clean
+                })
+            
+            i = j
+        else:
+            i += 1
+    
+    return instantiations
+
+def main():
+    # Device operations we're interested in (from previous analysis)
+    target_device_ops = [
+        'DeviceConv2dFwdXdl_C_Shuffle_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K',
+        'DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K',
+        'DeviceConvFwd',
+        'DeviceGroupedConvFwdDlMultipleD_NHWC_KYXC_NHWK',
+        'DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle',
+        'DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3',
+        'DeviceGroupedConvFwdMultipleD_Wmma_CShuffle',
+        'DeviceGroupedConvFwdMultipleD_Xdl_CShuffle_Large_Tensor',
+    ]
+    
+    # Define paths to search
+    header_paths = [
+        'library/include/ck/library/tensor_operation_instance/gpu/grouped_conv_fwd',
+        'library/include/ck/library/tensor_operation_instance/gpu',
+    ]
+    
+    source_paths = [
+        'library/src/tensor_operation_instance/gpu/grouped_conv1d_fwd',
+        'library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd',
+        'library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd',
+        'library/src/tensor_operation_instance/gpu/conv2d_fwd',
+    ]
+    
+    # Find all relevant files
+    header_files = find_files(header_paths, ['.hpp', '.h', '.inc'])
+    source_files = find_files(source_paths, ['.cpp', '.hpp', '.inc', '.in'])
+    
+    all_files = list(set(header_files + source_files))
+    
+    print(f"Analyzing {len(all_files)} files for detailed instantiations...\n")
+    
+    # Process each device operation
+    results = {}
+    
+    for device_op in target_device_ops:
+        print(f"Processing {device_op}...")
+        
+        device_op_data = {
+            'name': device_op,
+            'files_with_instantiations': {},
+            'total_instantiations': 0
+        }
+        
+        for file_path in all_files:
+            try:
+                with open(file_path, 'r', encoding='utf-8', errors='ignore') as f:
+                    content = f.read()
+                
+                if device_op in content:
+                    instantiations = extract_device_op_instantiations(content, device_op)
+                    
+                    if instantiations:
+                        rel_path = str(file_path).replace(os.getcwd() + '/', '')
+                        device_op_data['files_with_instantiations'][rel_path] = instantiations
+                        device_op_data['total_instantiations'] += len(instantiations)
+            except Exception as e:
+                print(f"  Error processing {file_path}: {e}")
+        
+        results[device_op] = device_op_data
+        print(f"  Found {device_op_data['total_instantiations']} instantiations in {len(device_op_data['files_with_instantiations'])} files")
+    
+    # Generate detailed report
+    report = []
+    report.append("=" * 100)
+    report.append("DETAILED FORWARD CONVOLUTION DEVICE OPERATIONS AND TEMPLATE INSTANTIATIONS")
+    report.append("=" * 100)
+    report.append("")
+    report.append(f"Generated for: ck::tensor_operation::device namespace")
+    report.append("")
+    
+    # Summary
+    report.append("SUMMARY")
+    report.append("-" * 100)
+    total_all_instantiations = sum(data['total_instantiations'] for data in results.values())
+    report.append(f"Total Device Operation Types: {len(target_device_ops)}")
+    report.append(f"Total Template Instantiations: {total_all_instantiations}")
+    report.append("")
+    
+    for device_op in target_device_ops:
+        data = results[device_op]
+        report.append(f"  • {device_op}")
+        report.append(f"    - Instantiations: {data['total_instantiations']}")
+        report.append(f"    - Files: {len(data['files_with_instantiations'])}")
+    
+    report.append("")
+    report.append("=" * 100)
+    report.append("")
+    
+    # Detailed section for each device op
+    for device_op in target_device_ops:
+        data = results[device_op]
+        
+        report.append("=" * 100)
+        report.append(f"DEVICE OPERATION: {device_op}")
+        report.append("=" * 100)
+        report.append("")
+        report.append(f"Total Template Instantiations: {data['total_instantiations']}")
+        report.append(f"Number of Files: {len(data['files_with_instantiations'])}")
+        report.append("")
+        
+        # List files and show sample instantiations
+        for file_path, instantiations in sorted(data['files_with_instantiations'].items()):
+            report.append("-" * 100)
+            report.append(f"File: {file_path}")
+            report.append(f"Instantiations in this file: {len(instantiations)}")
+            report.append("")
+            
+            # Show up to 3 sample instantiations from each file
+            num_samples = min(3, len(instantiations))
+            if num_samples > 0:
+                report.append("Sample instantiations:")
+                for idx, inst in enumerate(instantiations[:num_samples]):
+                    report.append(f"  [{idx+1}] Line {inst['line_start']}:")
+                    # Truncate very long instantiations
+                    text = inst['text']
+                    if len(text) > 500:
+                        text = text[:500] + "..."
+                    report.append(f"      {text}")
+                
+                if len(instantiations) > num_samples:
+                    report.append(f"  ... and {len(instantiations) - num_samples} more instantiations")
+            
+            report.append("")
+        
+        report.append("")
+    
+    # Save detailed report
+    report_text = '\n'.join(report)
+    
+    output_file = 'forward_conv_device_ops_detailed_report.txt'
+    with open(output_file, 'w') as f:
+        f.write(report_text)
+    
+    print(f"\n{'='*60}")
+    print(f"Detailed report saved to: {output_file}")
+    print(f"Total instantiations found: {total_all_instantiations}")
+    print(f"{'='*60}")
+
+if __name__ == '__main__':
+    main()

forward_conv_device_ops_detailed_report.txt

@@ -0,0 +1,428 @@
+====================================================================================================
+DETAILED FORWARD CONVOLUTION DEVICE OPERATIONS AND TEMPLATE INSTANTIATIONS
+====================================================================================================
+
+Generated for: ck::tensor_operation::device namespace
+
+SUMMARY
+----------------------------------------------------------------------------------------------------
+Total Device Operation Types: 8
+Total Template Instantiations: 1000
+
+  • DeviceConv2dFwdXdl_C_Shuffle_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K
+    - Instantiations: 61
+    - Files: 1
+  • DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K
+    - Instantiations: 156
+    - Files: 4
+  • DeviceConvFwd
+    - Instantiations: 12
+    - Files: 6
+  • DeviceGroupedConvFwdDlMultipleD_NHWC_KYXC_NHWK
+    - Instantiations: 6
+    - Files: 1
+  • DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle
+    - Instantiations: 570
+    - Files: 9
+  • DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3
+    - Instantiations: 141
+    - Files: 2
+  • DeviceGroupedConvFwdMultipleD_Wmma_CShuffle
+    - Instantiations: 42
+    - Files: 1
+  • DeviceGroupedConvFwdMultipleD_Xdl_CShuffle_Large_Tensor
+    - Instantiations: 12
+    - Files: 1
+
+====================================================================================================
+
+====================================================================================================
+DEVICE OPERATION: DeviceConv2dFwdXdl_C_Shuffle_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K
+====================================================================================================
+
+Total Template Instantiations: 61
+Number of Files: 1
+
+----------------------------------------------------------------------------------------------------
+File: library/src/tensor_operation_instance/gpu/conv2d_fwd/device_conv2d_fwd_xdl_c_shuffle_nhwc_kyxc_nhwk_f16_instance.cpp
+Instantiations in this file: 61
+
+Sample instantiations:
+  [1] Line 49:
+      DeviceConv2dFwdXdl_C_Shuffle_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough, ConvFwdDefault,   256,   256,   128,     4,  8,   32,   32,    4,    2,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,           1,           1,             S<1, 1, 32, 1, 1, 8>,  ...
+  [2] Line 50:
+      DeviceConv2dFwdXdl_C_Shuffle_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough, ConvFwdDefault,   256,   128,   256,     4,  8,   32,   32,    2,    4,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,           1,           1,             S<1, 1, 32, 1, 1, 8>,  ...
+  [3] Line 51:
+      DeviceConv2dFwdXdl_C_Shuffle_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough, ConvFwdDefault,   128,   128,   128,     4,  8,   32,   32,    4,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,           1,           1,             S<1, 1, 16, 1, 1, 8>,  ...
+  ... and 58 more instantiations
+
+
+====================================================================================================
+DEVICE OPERATION: DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K
+====================================================================================================
+
+Total Template Instantiations: 156
+Number of Files: 4
+
+----------------------------------------------------------------------------------------------------
+File: library/src/tensor_operation_instance/gpu/conv2d_fwd/device_conv2d_fwd_xdl_nhwc_kyxc_nhwk_bf16_instance.cpp
+Instantiations in this file: 39
+
+Sample instantiations:
+  [1] Line 45:
+      DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<   BF16,    BF16,    BF16,     F32, PassThrough, PassThrough, PassThrough, ConvFwdDefault,   256,   256,   128,     4,  8,   32,   32,    4,    2,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,               7,               1>,
+  [2] Line 46:
+      DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<   BF16,    BF16,    BF16,     F32, PassThrough, PassThrough, PassThrough, ConvFwdDefault,   256,   128,   256,     4,  8,   32,   32,    2,    4,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,               7,               1>,
+  [3] Line 47:
+      DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<   BF16,    BF16,    BF16,     F32, PassThrough, PassThrough, PassThrough, ConvFwdDefault,   128,   128,   128,     4,  8,   32,   32,    4,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,               7,               1>,
+  ... and 36 more instantiations
+
+----------------------------------------------------------------------------------------------------
+File: library/src/tensor_operation_instance/gpu/conv2d_fwd/device_conv2d_fwd_xdl_nhwc_kyxc_nhwk_f16_instance.cpp
+Instantiations in this file: 39
+
+Sample instantiations:
+  [1] Line 45:
+      DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough, ConvFwdDefault,   256,   256,   128,     4,  8,   32,   32,    4,    2,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,               7,               1>,
+  [2] Line 46:
+      DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough, ConvFwdDefault,   256,   128,   256,     4,  8,   32,   32,    2,    4,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,               7,               1>,
+  [3] Line 47:
+      DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough, ConvFwdDefault,   128,   128,   128,     4,  8,   32,   32,    4,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,               7,               1>,
+  ... and 36 more instantiations
+
+----------------------------------------------------------------------------------------------------
+File: library/src/tensor_operation_instance/gpu/conv2d_fwd/device_conv2d_fwd_xdl_nhwc_kyxc_nhwk_f32_instance.cpp
+Instantiations in this file: 39
+
+Sample instantiations:
+  [1] Line 44:
+      DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F32,     F32,     F32,     F32, PassThrough, PassThrough, PassThrough, ConvFwdDefault,   256,   256,   128,     4,  4,   32,   32,    4,    2,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              4,              4,      true,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              4,              4,      true,               7,               1>,
+  [2] Line 45:
+      DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F32,     F32,     F32,     F32, PassThrough, PassThrough, PassThrough, ConvFwdDefault,   256,   128,   256,     4,  4,   32,   32,    2,    4,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              4,              4,      true,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              4,              4,      true,               7,               1>,
+  [3] Line 46:
+      DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F32,     F32,     F32,     F32, PassThrough, PassThrough, PassThrough, ConvFwdDefault,   128,   128,   128,     4,  4,   32,   32,    4,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              4,              4,      true,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              4,              4,      true,               7,               1>,
+  ... and 36 more instantiations
+
+----------------------------------------------------------------------------------------------------
+File: library/src/tensor_operation_instance/gpu/conv2d_fwd/device_conv2d_fwd_xdl_nhwc_kyxc_nhwk_int8_instance.cpp
+Instantiations in this file: 39
+
+Sample instantiations:
+  [1] Line 42:
+      DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<  int8_t,  int8_t,  int8_t,  int32_t, PassThrough, PassThrough, PassThrough, ConvFwdDefault,   256,   256,   128,     4,  16,   32,   32,    4,    2,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,             16,             16,      true,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,             16,             16,      true,               7,               1>,
+  [2] Line 43:
+      DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<  int8_t,  int8_t,  int8_t,  int32_t, PassThrough, PassThrough, PassThrough, ConvFwdDefault,   256,   128,   256,     4,  16,   32,   32,    2,    4,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,             16,             16,      true,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,             16,             16,      true,               7,               1>,
+  [3] Line 44:
+      DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<  int8_t,  int8_t,  int8_t,  int32_t, PassThrough, PassThrough, PassThrough, ConvFwdDefault,   128,   128,   128,     4,  16,   32,   32,    4,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,             16,             16,      true,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,             16,             16,      true,               7,               1>,
+  ... and 36 more instantiations
+
+
+====================================================================================================
+DEVICE OPERATION: DeviceConvFwd
+====================================================================================================
+
+Total Template Instantiations: 12
+Number of Files: 6
+
+----------------------------------------------------------------------------------------------------
+File: library/include/ck/library/tensor_operation_instance/gpu/convolution_forward.hpp
+Instantiations in this file: 7
+
+Sample instantiations:
+  [1] Line 24:
+      DeviceConvFwd<2, NHWC, KYXC, NHWK, F16, F16, F16, PassThrough, PassThrough, PassThrough>>>&
+  [2] Line 28:
+      DeviceConvFwd<2, NHWC, KYXC, NHWK, F16, F16, F16, PassThrough, PassThrough, PassThrough>>>&
+  [3] Line 33:
+      std::vector<std::unique_ptr<DeviceConvFwd<2,
+                                              NHWC,
+                                              KYXC,
+                                              NHWK,
+                                              BF16,
+                                              BF16,
+                                              BF16,
+                                              PassThrough,
+                                              PassThrough,
+                         ...
+  ... and 4 more instantiations
+
+----------------------------------------------------------------------------------------------------
+File: library/src/tensor_operation_instance/gpu/conv2d_fwd/device_conv2d_fwd_xdl_c_shuffle_nhwc_kyxc_nhwk_f16_instance.cpp
+Instantiations in this file: 1
+
+Sample instantiations:
+  [1] Line 180:
+      DeviceConvFwd<2, NHWC, KYXC, NHWK, F16, F16, F16, PassThrough, PassThrough, PassThrough>>>&
+
+----------------------------------------------------------------------------------------------------
+File: library/src/tensor_operation_instance/gpu/conv2d_fwd/device_conv2d_fwd_xdl_nhwc_kyxc_nhwk_bf16_instance.cpp
+Instantiations in this file: 1
+
+Sample instantiations:
+  [1] Line 106:
+      std::vector<std::unique_ptr<DeviceConvFwd<2,
+                                              NHWC,
+                                              KYXC,
+                                              NHWK,
+                                              BF16,
+                                              BF16,
+                                              BF16,
+                                              PassThrough,
+                                              PassThrough,
+                         ...
+
+----------------------------------------------------------------------------------------------------
+File: library/src/tensor_operation_instance/gpu/conv2d_fwd/device_conv2d_fwd_xdl_nhwc_kyxc_nhwk_f16_instance.cpp
+Instantiations in this file: 1
+
+Sample instantiations:
+  [1] Line 107:
+      DeviceConvFwd<2, NHWC, KYXC, NHWK, F16, F16, F16, PassThrough, PassThrough, PassThrough>>>&
+
+----------------------------------------------------------------------------------------------------
+File: library/src/tensor_operation_instance/gpu/conv2d_fwd/device_conv2d_fwd_xdl_nhwc_kyxc_nhwk_f32_instance.cpp
+Instantiations in this file: 1
+
+Sample instantiations:
+  [1] Line 106:
+      DeviceConvFwd<2, NHWC, KYXC, NHWK, F32, F32, F32, PassThrough, PassThrough, PassThrough>>>&
+
+----------------------------------------------------------------------------------------------------
+File: library/src/tensor_operation_instance/gpu/conv2d_fwd/device_conv2d_fwd_xdl_nhwc_kyxc_nhwk_int8_instance.cpp
+Instantiations in this file: 1
+
+Sample instantiations:
+  [1] Line 103:
+      std::vector<std::unique_ptr<DeviceConvFwd<2,
+                                              NHWC,
+                                              KYXC,
+                                              NHWK,
+                                              int8_t,
+                                              int8_t,
+                                              int8_t,
+                                              PassThrough,
+                                              PassThrough,
+                   ...
+
+
+====================================================================================================
+DEVICE OPERATION: DeviceGroupedConvFwdDlMultipleD_NHWC_KYXC_NHWK
+====================================================================================================
+
+Total Template Instantiations: 6
+Number of Files: 1
+
+----------------------------------------------------------------------------------------------------
+File: library/include/ck/library/tensor_operation_instance/gpu/grouped_conv_fwd/device_grouped_conv_fwd_dl_instance.hpp
+Instantiations in this file: 6
+
+Sample instantiations:
+  [1] Line 57:
+      DeviceGroupedConvFwdDlMultipleD_NHWC_KYXC_NHWK<           2,    F16,     F16,  DsDatatype,     F16,     F32, InLayout, WeiLayout,  DsLayout, OutLayout, PassThrough,  PassThrough,  CDEElementOp,       ConvSpec,    GemmMNKPadding,     8,    16,     4,     2,  1,          1,          2,      1,       S<4, 2>,       S<1, 1>,      S<2, 1, 2, 1>,        S<1, 1, 8, 1>,  S<1, 2, 0, 3>,  S<1, 2, 0, 3>,       S<1, 1, 1, 1>,      S<1, 2, 0, 3>,       S<1, 1, 1, 1>,      S<1, 1, 1, 1>,        S<2, 1, 4, 1>,...
+  [2] Line 58:
+      DeviceGroupedConvFwdDlMultipleD_NHWC_KYXC_NHWK<           2,    F16,     F16,  DsDatatype,     F16,     F32, InLayout, WeiLayout,  DsLayout, OutLayout, PassThrough,  PassThrough,  CDEElementOp,       ConvSpec,    GemmMNKPadding,   256,   128,   128,    16,  1,          4,          4,      1,       S<8, 2>,       S<8, 2>,      S<8, 1, 1, 1>,      S<2, 1, 128, 1>,  S<1, 2, 0, 3>,  S<1, 2, 0, 3>,       S<1, 1, 1, 1>,      S<1, 2, 0, 3>,       S<1, 1, 1, 1>,      S<8, 1, 1, 1>,      S<2, 1, 128, 1>,...
+  [3] Line 60:
+      DeviceGroupedConvFwdDlMultipleD_NHWC_KYXC_NHWK<           2,    F16,     F16,  DsDatatype,     F16,     F32, InLayout, WeiLayout,  DsLayout, OutLayout, PassThrough,  PassThrough,  CDEElementOp,       ConvSpec,    GemmMNKPadding,   256,   128,   128,    16,  2,          4,          4,      1,       S<8, 2>,       S<8, 2>,      S<8, 1, 1, 2>,      S<2, 1, 128, 1>,  S<1, 2, 0, 3>,  S<1, 2, 0, 3>,       S<4, 1, 1, 2>,      S<1, 2, 0, 3>,       S<1, 1, 1, 2>,      S<8, 1, 1, 2>,      S<2, 1, 128, 1>,...
+  ... and 3 more instantiations
+
+
+====================================================================================================
+DEVICE OPERATION: DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle
+====================================================================================================
+
+Total Template Instantiations: 570
+Number of Files: 9
+
+----------------------------------------------------------------------------------------------------
+File: library/include/ck/library/tensor_operation_instance/gpu/grouped_conv_fwd/device_grouped_conv_fwd_xdl_bilinear_instance.hpp
+Instantiations in this file: 80
+
+Sample instantiations:
+  [1] Line 54:
+      DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<NDimSpatial,ALayout,BLayout,    DsLayout,ELayout,  BF16,  BF16,     F32,     BF16, Tuple<BF16>,  BF16, PassThrough, PassThrough,  Bilinear,       ConvSpec, GemmMNKPadding,        1,    64,    64,    64,    32,   8,   8,   32,   32,    2,    2,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              1,              8,         1,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              1,              8,        ...
+  [2] Line 56:
+      DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<NDimSpatial,ALayout,BLayout,    DsLayout,ELayout,  BF16,  BF16,     F32,     BF16, Tuple<BF16>,  BF16, PassThrough, PassThrough,  Bilinear,       ConvSpec, GemmMNKPadding,        1,    64,    64,    32,    32,   8,   8,   32,   32,    2,    1,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,        ...
+  [3] Line 57:
+      DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<NDimSpatial,ALayout,BLayout,    DsLayout,ELayout,  BF16,  BF16,     F32,     BF16, Tuple<BF16>,  BF16, PassThrough, PassThrough,  Bilinear,       ConvSpec, GemmMNKPadding,        1,   256,   128,   128,    32,   8,   8,   32,   32,    2,    2,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              1,              8,         1,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              1,              8,        ...
+  ... and 77 more instantiations
+
+----------------------------------------------------------------------------------------------------
+File: library/include/ck/library/tensor_operation_instance/gpu/grouped_conv_fwd/device_grouped_conv_fwd_xdl_binary_outelementop_instance.hpp
+Instantiations in this file: 16
+
+Sample instantiations:
+  [1] Line 53:
+      DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<NDimSpatial,ALayout,BLayout,  DsLayout,ELayout,    F8,    F8,     F32,      F32,  Tuple<F32>,    F8, PassThrough, PassThrough,  OutElementOp,       ConvSpec, GemmMNKPadding,        1,    64,    64,    64,    32,   8,   8,   32,   32,    2,    2,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              1,              8,         1,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              1,              8,      ...
+  [2] Line 55:
+      DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<NDimSpatial,ALayout,BLayout,  DsLayout,ELayout,    F8,    F8,     F32,      F32,  Tuple<F32>,    F8, PassThrough, PassThrough,  OutElementOp,       ConvSpec, GemmMNKPadding,        1,    64,    64,    32,    32,   8,   8,   32,   32,    2,    1,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      ...
+  [3] Line 56:
+      DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<NDimSpatial,ALayout,BLayout,  DsLayout,ELayout,    F8,    F8,     F32,      F32,  Tuple<F32>,    F8, PassThrough, PassThrough,  OutElementOp,       ConvSpec, GemmMNKPadding,        1,   256,   128,   128,    32,   8,   8,   32,   32,    2,    2,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              1,              8,         1,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              1,              8,      ...
+  ... and 13 more instantiations
+
+----------------------------------------------------------------------------------------------------
+File: library/include/ck/library/tensor_operation_instance/gpu/grouped_conv_fwd/device_grouped_conv_fwd_xdl_dynamic_op_instance.hpp
+Instantiations in this file: 64
+
+Sample instantiations:
+  [1] Line 54:
+      DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<NDimSpatial,ALayout,BLayout,    DsLayout,ELayout,  BF16,  BF16,     F32,     BF16,    Tuple<>,  BF16, PassThrough, PassThrough,     DynamicUnaryOp,       ConvSpec, GemmMNKPadding,        1,    64,    64,    64,    32,   8,   8,   32,   32,    2,    2,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              1,              8,         1,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              1,              8,...
+  [2] Line 56:
+      DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<NDimSpatial,ALayout,BLayout,    DsLayout,ELayout,  BF16,  BF16,     F32,     BF16,    Tuple<>,  BF16, PassThrough, PassThrough,     DynamicUnaryOp,       ConvSpec, GemmMNKPadding,        1,    64,    64,    32,    32,   8,   8,   32,   32,    2,    1,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,...
+  [3] Line 57:
+      DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<NDimSpatial,ALayout,BLayout,    DsLayout,ELayout,  BF16,  BF16,     F32,     BF16,    Tuple<>,  BF16, PassThrough, PassThrough,     DynamicUnaryOp,       ConvSpec, GemmMNKPadding,        1,   256,   128,   128,    32,   8,   8,   32,   32,    2,    2,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              1,              8,         1,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              1,              8,...
+  ... and 61 more instantiations
+
+----------------------------------------------------------------------------------------------------
+File: library/include/ck/library/tensor_operation_instance/gpu/grouped_conv_fwd/device_grouped_conv_fwd_xdl_instance.hpp
+Instantiations in this file: 193
+
+Sample instantiations:
+  [1] Line 68:
+      DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<NDimSpatial,ALayout,BLayout,    DsLayout,ELayout,  BF16,  BF16,     F32,     BF16,    DsDataTypes,  BF16, PassThrough, PassThrough, OutElementOp,       ConvSpec, GemmMNKPadding,        1,    64,    64,    64,    32,   8,   8,   32,   32,    2,    2,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              1,              8,         1,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              1,              8,  ...
+  [2] Line 87:
+      DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<NDimSpatial,ALayout,BLayout,    DsLayout,ELayout,  BF16,  BF16,     F32,     BF16,    DsDataTypes,  BF16, PassThrough, PassThrough, OutElementOp,       ConvSpec, GemmMNKPadding,        1,    64,    64,    64,    32,   8,   8,   32,   32,    2,    2,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              1,              8,         1,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              1,              8,  ...
+  [3] Line 89:
+      DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<NDimSpatial,ALayout,BLayout,    DsLayout,ELayout,  BF16,  BF16,     F32,     BF16,    DsDataTypes,  BF16, PassThrough, PassThrough, OutElementOp,       ConvSpec, GemmMNKPadding,        1,    64,    64,    32,    32,   8,   8,   32,   32,    2,    1,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,  ...
+  ... and 190 more instantiations
+
+----------------------------------------------------------------------------------------------------
+File: library/include/ck/library/tensor_operation_instance/gpu/grouped_conv_fwd/device_grouped_conv_fwd_xdl_merged_groups_instance.hpp
+Instantiations in this file: 21
+
+Sample instantiations:
+  [1] Line 55:
+      DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<NDimSpatial,ALayout,BLayout,    DsLayout,ELayout,  BF16,  BF16,     F32,     BF16,    DsDataTypes,  BF16, PassThrough, PassThrough, OutElementOp,                  ConvSpec, GemmMNKPadding,  1,  64,    64,    16,     16,   4, 4,  16,   16,    4,    1,  S< 4, 16,  1>, S<0, 2, 1>,     S<0, 2, 1>,                   1,              4,              4,      1,  S< 4, 16,  1>,   S<1, 0, 2>,     S<1, 0, 2>,             2,              1,              4,      1...
+  [2] Line 56:
+      DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<NDimSpatial,ALayout,BLayout,    DsLayout,ELayout,  BF16,  BF16,     F32,     BF16,    DsDataTypes,  BF16, PassThrough, PassThrough, OutElementOp,                  ConvSpec, GemmMNKPadding,  1,  64,    64,    16,     16,   4, 4,  16,   16,    4,    1,  S< 4, 16,  1>, S<0, 2, 1>,     S<0, 2, 1>,                   1,              4,              4,      1,  S< 4, 16,  1>,   S<1, 0, 2>,     S<1, 0, 2>,             2,              1,              4,      1...
+  [3] Line 57:
+      DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<NDimSpatial,ALayout,BLayout,    DsLayout,ELayout,  BF16,  BF16,     F32,     BF16,    DsDataTypes,  BF16, PassThrough, PassThrough, OutElementOp,                  ConvSpec, GemmMNKPadding,  1,  64,    64,    16,     16,   4, 4,  16,   16,    4,    1,  S< 4, 16,  1>, S<0, 2, 1>,     S<0, 2, 1>,                   1,              4,              4,      1,  S< 4, 16,  1>,   S<1, 0, 2>,     S<1, 0, 2>,             2,              1,              4,      1...
+  ... and 18 more instantiations
+
+----------------------------------------------------------------------------------------------------
+File: library/include/ck/library/tensor_operation_instance/gpu/grouped_conv_fwd/device_grouped_conv_fwd_xdl_outelementop_instance.hpp
+Instantiations in this file: 80
+
+Sample instantiations:
+  [1] Line 54:
+      DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<NDimSpatial,ALayout,BLayout,  DsLayout,ELayout,    F8,    F8,     F32,      F32,  Tuple<>,    F8, PassThrough, PassThrough,  OutElementOp,       ConvSpec, GemmMNKPadding,        1,    64,    64,    64,    32,   8,   8,   32,   32,    2,    2,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              1,              8,         1,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              1,              8,         ...
+  [2] Line 56:
+      DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<NDimSpatial,ALayout,BLayout,  DsLayout,ELayout,    F8,    F8,     F32,      F32,  Tuple<>,    F8, PassThrough, PassThrough,  OutElementOp,       ConvSpec, GemmMNKPadding,        1,    64,    64,    32,    32,   8,   8,   32,   32,    2,    1,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         ...
+  [3] Line 57:
+      DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<NDimSpatial,ALayout,BLayout,  DsLayout,ELayout,    F8,    F8,     F32,      F32,  Tuple<>,    F8, PassThrough, PassThrough,  OutElementOp,       ConvSpec, GemmMNKPadding,        1,   256,   128,   128,    32,   8,   8,   32,   32,    2,    2,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              1,              8,         1,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              1,              8,         ...
+  ... and 77 more instantiations
+
+----------------------------------------------------------------------------------------------------
+File: library/include/ck/library/tensor_operation_instance/gpu/grouped_conv_fwd/device_grouped_conv_fwd_xdl_scale_instance.hpp
+Instantiations in this file: 80
+
+Sample instantiations:
+  [1] Line 54:
+      DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<NDimSpatial,ALayout,BLayout,    DsLayout,ELayout,  BF16,  BF16,     F32,     BF16,    Tuple<>,  BF16, PassThrough, PassThrough,     Scale,       ConvSpec, GemmMNKPadding,        1,    64,    64,    64,    32,   8,   8,   32,   32,    2,    2,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              1,              8,         1,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              1,              8,         ...
+  [2] Line 56:
+      DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<NDimSpatial,ALayout,BLayout,    DsLayout,ELayout,  BF16,  BF16,     F32,     BF16,    Tuple<>,  BF16, PassThrough, PassThrough,     Scale,       ConvSpec, GemmMNKPadding,        1,    64,    64,    32,    32,   8,   8,   32,   32,    2,    1,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         ...
+  [3] Line 57:
+      DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<NDimSpatial,ALayout,BLayout,    DsLayout,ELayout,  BF16,  BF16,     F32,     BF16,    Tuple<>,  BF16, PassThrough, PassThrough,     Scale,       ConvSpec, GemmMNKPadding,        1,   256,   128,   128,    32,   8,   8,   32,   32,    2,    2,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              1,              8,         1,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              1,              8,         ...
+  ... and 77 more instantiations
+
+----------------------------------------------------------------------------------------------------
+File: library/include/ck/library/tensor_operation_instance/gpu/grouped_conv_fwd/device_grouped_conv_fwd_xdl_scaleadd_ab_instance.hpp
+Instantiations in this file: 20
+
+Sample instantiations:
+  [1] Line 50:
+      DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<NDimSpatial,ALayout,BLayout, ck::Tuple<>,ELayout,  ck::Tuple<BF16,  BF16>, ck::Tuple<BF16,  BF16>,     F32, BF16, ck::Tuple<>, BF16,  ScaleAdd, ScaleAdd, PassThrough,         ConvSpec, GemmMNKPadding,        1,    64,    64,    64,    32,   8,   8,   32,   32,    2,    2,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              1,              1,         1,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,           ...
+  [2] Line 52:
+      DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<NDimSpatial,ALayout,BLayout, ck::Tuple<>,ELayout,  ck::Tuple<BF16,  BF16>, ck::Tuple<BF16,  BF16>,     F32, BF16, ck::Tuple<>, BF16,  ScaleAdd, ScaleAdd, PassThrough,         ConvSpec, GemmMNKPadding,        1,    64,    64,    32,    32,   8,   8,   32,   32,    2,    1,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,           ...
+  [3] Line 53:
+      DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<NDimSpatial,ALayout,BLayout, ck::Tuple<>,ELayout,  ck::Tuple<BF16,  BF16>, ck::Tuple<BF16,  BF16>,     F32, BF16, ck::Tuple<>, BF16,  ScaleAdd, ScaleAdd, PassThrough,         ConvSpec, GemmMNKPadding,        1,   256,   128,   128,    32,   8,   8,   32,   32,    2,    2,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              1,              1,         1,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,           ...
+  ... and 17 more instantiations
+
+----------------------------------------------------------------------------------------------------
+File: library/include/ck/library/tensor_operation_instance/gpu/grouped_conv_fwd/device_grouped_conv_fwd_xdl_scaleadd_scaleadd_relu_instance.hpp
+Instantiations in this file: 16
+
+Sample instantiations:
+  [1] Line 53:
+      DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<NDimSpatial,ALayout,BLayout,    DsLayout,ELayout,  BF16,  BF16,     F32, BF16, ck::Tuple<BF16, BF16>, BF16, PassThrough, PassThrough, ScaleAddScaleAddRelu,  ConvSpec, GemmMNKPadding,        1,    64,    64,    64,    32,   8,   8,   32,   32,    2,    2,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              1,              8,         1,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              1,             ...
+  [2] Line 55:
+      DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<NDimSpatial,ALayout,BLayout,    DsLayout,ELayout,  BF16,  BF16,     F32, BF16, ck::Tuple<BF16, BF16>, BF16, PassThrough, PassThrough, ScaleAddScaleAddRelu,  ConvSpec, GemmMNKPadding,        1,    64,    64,    32,    32,   8,   8,   32,   32,    2,    1,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,             ...
+  [3] Line 56:
+      DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<NDimSpatial,ALayout,BLayout,    DsLayout,ELayout,  BF16,  BF16,     F32, BF16, ck::Tuple<BF16, BF16>, BF16, PassThrough, PassThrough, ScaleAddScaleAddRelu,  ConvSpec, GemmMNKPadding,        1,   256,   128,   128,    32,   8,   8,   32,   32,    2,    2,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              1,              8,         1,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              1,             ...
+  ... and 13 more instantiations
+
+
+====================================================================================================
+DEVICE OPERATION: DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3
+====================================================================================================
+
+Total Template Instantiations: 141
+Number of Files: 2
+
+----------------------------------------------------------------------------------------------------
+File: library/include/ck/library/tensor_operation_instance/gpu/grouped_conv_fwd/device_grouped_conv_fwd_xdl_comp_instance.hpp
+Instantiations in this file: 49
+
+Sample instantiations:
+  [1] Line 71:
+      DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<NDimSpatial,ALayout,BLayout,    DsLayout,ELayout,  BF16,  BF16,     F32,     BF16,    DsDataTypes,  BF16, PassThrough, PassThrough, OutElementOp,       ConvSpec, GemmMNKPadding,   256,   128,   128,    64,  16,  16,  32,   32,    2,    2,     S<4, 64, 1>,     S<1, 0, 2>,    S<1, 0, 2>,               2,              8,              8,          1,    S<4, 64, 1>,     S<1, 0, 2>,    S<1, 0, 2>,               2,              8,              8,         ...
+  [2] Line 90:
+      DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<NDimSpatial,ALayout,BLayout,    DsLayout,ELayout,  BF16,  BF16,     F32,     BF16,    DsDataTypes,  BF16, PassThrough, PassThrough, OutElementOp,       ConvSpec, GemmMNKPadding,   256,   256,   256,    32,   8,   8,  32,   32,    4,    4,     S<4, 64, 1>,     S<1, 0, 2>,    S<1, 0, 2>,               2,              8,              8,          0,    S<4, 64, 1>,     S<1, 0, 2>,    S<1, 0, 2>,               2,              8,              8,         ...
+  [3] Line 91:
+      DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<NDimSpatial,ALayout,BLayout,    DsLayout,ELayout,  BF16,  BF16,     F32,     BF16,    DsDataTypes,  BF16, PassThrough, PassThrough, OutElementOp,       ConvSpec, GemmMNKPadding,   256,   128,   128,    64,   8,   8,  32,   32,    2,    2,     S<8, 32, 1>,     S<1, 0, 2>,    S<1, 0, 2>,               2,              8,              8,          0,    S<8, 32, 1>,     S<1, 0, 2>,    S<1, 0, 2>,               2,              8,              8,         ...
+  ... and 46 more instantiations
+
+----------------------------------------------------------------------------------------------------
+File: library/include/ck/library/tensor_operation_instance/gpu/grouped_conv_fwd/device_grouped_conv_fwd_xdl_mem_instance.hpp
+Instantiations in this file: 92
+
+Sample instantiations:
+  [1] Line 69:
+      DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<NDimSpatial,ALayout,BLayout,    DsLayout,ELayout,  BF16,  BF16,     F32,     BF16,    DsDataTypes,  BF16, PassThrough, PassThrough, OutElementOp,       ConvSpec, GemmMNKPadding,   128,    32,   16,    64,   8,   8,  16,   16,    1,    1,     S<8, 16, 1>,     S<1, 0, 2>,    S<1, 0, 2>,               2,              8,              8,          0,    S<8, 16, 1>,     S<1, 0, 2>,    S<1, 0, 2>,               2,              8,              8,          ...
+  [2] Line 70:
+      DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<NDimSpatial,ALayout,BLayout,    DsLayout,ELayout,  BF16,  BF16,     F32,     BF16,    DsDataTypes,  BF16, PassThrough, PassThrough, OutElementOp,       ConvSpec, GemmMNKPadding,    64,    16,   16,   128,   8,   8,  16,   16,    1,    1,     S<16, 4, 1>,     S<1, 0, 2>,    S<1, 0, 2>,               2,              8,              8,          0,    S<16, 4, 1>,     S<1, 0, 2>,    S<1, 0, 2>,               2,              8,              8,          ...
+  [3] Line 71:
+      DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<NDimSpatial,ALayout,BLayout,    DsLayout,ELayout,  BF16,  BF16,     F32,     BF16,    DsDataTypes,  BF16, PassThrough, PassThrough, OutElementOp,       ConvSpec, GemmMNKPadding,    64,    16,   16,    64,   8,   8,  16,   16,    1,    1,     S<8,  8, 1>,     S<1, 0, 2>,    S<1, 0, 2>,               2,              8,              8,          0,    S<8,  8, 1>,     S<1, 0, 2>,    S<1, 0, 2>,               2,              8,              8,          ...
+  ... and 89 more instantiations
+
+
+====================================================================================================
+DEVICE OPERATION: DeviceGroupedConvFwdMultipleD_Wmma_CShuffle
+====================================================================================================
+
+Total Template Instantiations: 42
+Number of Files: 1
+
+----------------------------------------------------------------------------------------------------
+File: library/include/ck/library/tensor_operation_instance/gpu/grouped_conv_fwd/device_grouped_conv_fwd_wmma_instance.hpp
+Instantiations in this file: 42
+
+Sample instantiations:
+  [1] Line 62:
+      DeviceGroupedConvFwdMultipleD_Wmma_CShuffle<NDSpatial, ALayout, BLayout, DsLayout, ELayout,  F16,   F16,     F32,      F16, DsDatatype,    F16,  PassThrough, PassThrough, CDEElementOp,       ConvSpec, GemmMNKPadding,         1,    128,    64,    64,    32,  8,    16,   16,       2,       2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              1,              8,         1,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              1,              8,     ...
+  [2] Line 64:
+      DeviceGroupedConvFwdMultipleD_Wmma_CShuffle<NDSpatial, ALayout, BLayout, DsLayout, ELayout,  F16,   F16,     F32,      F16, DsDatatype,    F16,  PassThrough, PassThrough, CDEElementOp,       ConvSpec, GemmMNKPadding,         1,    256,   128,   128,    32,  8,    16,   16,       4,       2,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,     ...
+  [3] Line 65:
+      DeviceGroupedConvFwdMultipleD_Wmma_CShuffle<NDSpatial, ALayout, BLayout, DsLayout, ELayout,  F16,   F16,     F32,      F16, DsDatatype,    F16,  PassThrough, PassThrough, CDEElementOp,       ConvSpec, GemmMNKPadding,         1,    256,    64,   256,    32,  8,    16,   16,       2,       4,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,     ...
+  ... and 39 more instantiations
+
+
+====================================================================================================
+DEVICE OPERATION: DeviceGroupedConvFwdMultipleD_Xdl_CShuffle_Large_Tensor
+====================================================================================================
+
+Total Template Instantiations: 12
+Number of Files: 1
+
+----------------------------------------------------------------------------------------------------
+File: library/include/ck/library/tensor_operation_instance/gpu/grouped_conv_fwd/device_grouped_conv_fwd_xdl_large_tensor_instance.hpp
+Instantiations in this file: 12
+
+Sample instantiations:
+  [1] Line 53:
+      DeviceGroupedConvFwdMultipleD_Xdl_CShuffle_Large_Tensor<NDimSpatial,ALayout,BLayout,    DsLayout,ELayout,  BF16,  BF16,     F32,     BF16,    DsDataTypes,  BF16, PassThrough, PassThrough, OutElementOp,       ConvSpec, GemmMNKPadding,        1,    64,    64,    64,    32,   8,   8,   32,   32,    2,    2,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              1,              8,         1,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              1,       ...
+  [2] Line 55:
+      DeviceGroupedConvFwdMultipleD_Xdl_CShuffle_Large_Tensor<NDimSpatial,ALayout,BLayout,    DsLayout,ELayout,  BF16,  BF16,     F32,     BF16,    DsDataTypes,  BF16, PassThrough, PassThrough, OutElementOp,       ConvSpec, GemmMNKPadding,        1,   256,   256,   128,    32,   8,   8,   32,   32,    4,    2,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              2,              8,         1,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              2,       ...
+  [3] Line 56:
+      DeviceGroupedConvFwdMultipleD_Xdl_CShuffle_Large_Tensor<NDimSpatial,ALayout,BLayout,    DsLayout,ELayout,  BF16,  BF16,     F32,     BF16,    DsDataTypes,  BF16, PassThrough, PassThrough, OutElementOp,       ConvSpec, GemmMNKPadding,        1,   256,   256,   128,    32,   8,   8,   32,   32,    4,    2,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,       ...
+  ... and 9 more instantiations
+

forward_convolution_device_ops_data.json

@@ -0,0 +1,285 @@
+{
+  "summary": {
+    "total_device_operations": 9,
+    "device_operation_names": [
+      "DeviceConv2dFwdXdl_C_Shuffle_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K",
+      "DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K",
+      "DeviceConvFwd",
+      "DeviceGroupedConvFwdDlMultipleD_NHWC_KYXC_NHWK",
+      "DeviceGroupedConvFwdMultipleABD",
+      "DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle",
+      "DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3",
+      "DeviceGroupedConvFwdMultipleD_Wmma_CShuffle",
+      "DeviceGroupedConvFwdMultipleD_Xdl_CShuffle_Large_Tensor"
+    ]
+  },
+  "details": {
+    "DeviceConv2dFwdXdl_C_Shuffle_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K": {
+      "files": [
+        "library/src/tensor_operation_instance/gpu/conv2d_fwd/device_conv2d_fwd_xdl_c_shuffle_nhwc_kyxc_nhwk_f16_instance.cpp"
+      ],
+      "total_instantiations": 61
+    },
+    "DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K": {
+      "files": [
+        "library/src/tensor_operation_instance/gpu/conv2d_fwd/device_conv2d_fwd_xdl_nhwc_kyxc_nhwk_bf16_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/conv2d_fwd/device_conv2d_fwd_xdl_nhwc_kyxc_nhwk_f16_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/conv2d_fwd/device_conv2d_fwd_xdl_nhwc_kyxc_nhwk_f32_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/conv2d_fwd/device_conv2d_fwd_xdl_nhwc_kyxc_nhwk_int8_instance.cpp"
+      ],
+      "total_instantiations": 156
+    },
+    "DeviceConvFwd": {
+      "files": [
+        "library/include/ck/library/tensor_operation_instance/gpu/convolution_forward.hpp",
+        "library/src/tensor_operation_instance/gpu/conv2d_fwd/device_conv2d_fwd_xdl_c_shuffle_nhwc_kyxc_nhwk_f16_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/conv2d_fwd/device_conv2d_fwd_xdl_nhwc_kyxc_nhwk_bf16_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/conv2d_fwd/device_conv2d_fwd_xdl_nhwc_kyxc_nhwk_f16_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/conv2d_fwd/device_conv2d_fwd_xdl_nhwc_kyxc_nhwk_f32_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/conv2d_fwd/device_conv2d_fwd_xdl_nhwc_kyxc_nhwk_int8_instance.cpp"
+      ],
+      "total_instantiations": 12
+    },
+    "DeviceGroupedConvFwdDlMultipleD_NHWC_KYXC_NHWK": {
+      "files": [
+        "library/include/ck/library/tensor_operation_instance/gpu/grouped_conv_fwd/device_grouped_conv_fwd_dl_instance.hpp"
+      ],
+      "total_instantiations": 6
+    },
+    "DeviceGroupedConvFwdMultipleABD": {
+      "files": [
+        "library/include/ck/library/tensor_operation_instance/gpu/grouped_conv_fwd/device_grouped_conv_fwd_xdl_bilinear_instance.hpp",
+        "library/include/ck/library/tensor_operation_instance/gpu/grouped_conv_fwd/device_grouped_conv_fwd_xdl_binary_outelementop_instance.hpp",
+        "library/include/ck/library/tensor_operation_instance/gpu/grouped_conv_fwd/device_grouped_conv_fwd_xdl_comp_instance.hpp",
+        "library/include/ck/library/tensor_operation_instance/gpu/grouped_conv_fwd/device_grouped_conv_fwd_xdl_dynamic_op_instance.hpp",
+        "library/include/ck/library/tensor_operation_instance/gpu/grouped_conv_fwd/device_grouped_conv_fwd_xdl_instance.hpp",
+        "library/include/ck/library/tensor_operation_instance/gpu/grouped_conv_fwd/device_grouped_conv_fwd_xdl_mem_instance.hpp",
+        "library/include/ck/library/tensor_operation_instance/gpu/grouped_conv_fwd/device_grouped_conv_fwd_xdl_merged_groups_instance.hpp",
+        "library/include/ck/library/tensor_operation_instance/gpu/grouped_conv_fwd/device_grouped_conv_fwd_xdl_outelementop_instance.hpp",
+        "library/include/ck/library/tensor_operation_instance/gpu/grouped_conv_fwd/device_grouped_conv_fwd_xdl_scale_instance.hpp",
+        "library/include/ck/library/tensor_operation_instance/gpu/grouped_conv_fwd/device_grouped_conv_fwd_xdl_scaleadd_ab_instance.hpp",
+        "library/include/ck/library/tensor_operation_instance/gpu/grouped_conv_fwd/device_grouped_conv_fwd_xdl_scaleadd_scaleadd_relu_instance.hpp",
+        "library/include/ck/library/tensor_operation_instance/gpu/grouped_convolution_forward.hpp",
+        "library/include/ck/library/tensor_operation_instance/gpu/grouped_convolution_forward_bias_bnorm_clamp.hpp",
+        "library/include/ck/library/tensor_operation_instance/gpu/grouped_convolution_forward_bias_clamp.hpp",
+        "library/include/ck/library/tensor_operation_instance/gpu/grouped_convolution_forward_bilinear.hpp",
+        "library/include/ck/library/tensor_operation_instance/gpu/grouped_convolution_forward_clamp.hpp",
+        "library/include/ck/library/tensor_operation_instance/gpu/grouped_convolution_forward_convinvscale.hpp",
+        "library/include/ck/library/tensor_operation_instance/gpu/grouped_convolution_forward_convscale.hpp",
+        "library/include/ck/library/tensor_operation_instance/gpu/grouped_convolution_forward_convscale_add.hpp",
+        "library/include/ck/library/tensor_operation_instance/gpu/grouped_convolution_forward_convscale_relu.hpp",
+        "library/include/ck/library/tensor_operation_instance/gpu/grouped_convolution_forward_dynamic_op.hpp",
+        "library/include/ck/library/tensor_operation_instance/gpu/grouped_convolution_forward_scale.hpp",
+        "library/include/ck/library/tensor_operation_instance/gpu/grouped_convolution_forward_scaleadd_ab.hpp",
+        "library/include/ck/library/tensor_operation_instance/gpu/grouped_convolution_forward_scaleadd_scaleadd_relu.hpp",
+        "library/include/ck/library/tensor_operation_instance/gpu/quantization/grouped_convolution_bias_forward_perchannel_quantization.hpp",
+        "library/include/ck/library/tensor_operation_instance/gpu/quantization/grouped_convolution_bias_forward_perlayer_quantization.hpp",
+        "library/include/ck/library/tensor_operation_instance/gpu/quantization/grouped_convolution_forward_perchannel_quantization.hpp",
+        "library/include/ck/library/tensor_operation_instance/gpu/quantization/grouped_convolution_forward_perlayer_quantization.hpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv1d_fwd/xdl/device_grouped_conv1d_fwd_xdl_gnwc_gkxc_gnwk_bf16_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv1d_fwd/xdl/device_grouped_conv1d_fwd_xdl_gnwc_gkxc_gnwk_f16_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv1d_fwd/xdl/device_grouped_conv1d_fwd_xdl_gnwc_gkxc_gnwk_f32_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv1d_fwd/xdl/device_grouped_conv1d_fwd_xdl_gnwc_gkxc_gnwk_int8_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/dl/device_grouped_conv2d_fwd_dl_gnhwc_gkyxc_gnhwk_f16_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/dl/device_grouped_conv2d_fwd_dl_gnhwc_gkyxc_gnhwk_f32_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/dl/device_grouped_conv2d_fwd_dl_nhwgc_gkyxc_nhwgk_f16_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/dl/device_grouped_conv2d_fwd_dl_nhwgc_gkyxc_nhwgk_f32_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/wmma/device_grouped_conv2d_fwd_wmma_gnhwc_gkyxc_gnhwk_f16_1x1p0_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/wmma/device_grouped_conv2d_fwd_wmma_gnhwc_gkyxc_gnhwk_f16_1x1s1p0_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/wmma/device_grouped_conv2d_fwd_wmma_gnhwc_gkyxc_gnhwk_f16_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/wmma/device_grouped_conv2d_fwd_wmma_gnhwc_gkyxc_gnhwk_f16_oddc_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/wmma/device_grouped_conv2d_fwd_wmma_gnhwc_gkyxc_gnhwk_i8_1x1p0_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/wmma/device_grouped_conv2d_fwd_wmma_gnhwc_gkyxc_gnhwk_i8_1x1s1p0_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/wmma/device_grouped_conv2d_fwd_wmma_gnhwc_gkyxc_gnhwk_i8_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/wmma/device_grouped_conv2d_fwd_wmma_gnhwc_gkyxc_gnhwk_i8_oddc_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/wmma/device_grouped_conv2d_fwd_wmma_nhwgc_gkyxc_nhwgk_f16_1x1p0_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/wmma/device_grouped_conv2d_fwd_wmma_nhwgc_gkyxc_nhwgk_f16_1x1s1p0_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/wmma/device_grouped_conv2d_fwd_wmma_nhwgc_gkyxc_nhwgk_f16_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/wmma/device_grouped_conv2d_fwd_wmma_nhwgc_gkyxc_nhwgk_f16_oddc_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/wmma/device_grouped_conv2d_fwd_wmma_nhwgc_gkyxc_nhwgk_i8_1x1p0_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/wmma/device_grouped_conv2d_fwd_wmma_nhwgc_gkyxc_nhwgk_i8_1x1s1p0_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/wmma/device_grouped_conv2d_fwd_wmma_nhwgc_gkyxc_nhwgk_i8_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/wmma/device_grouped_conv2d_fwd_wmma_nhwgc_gkyxc_nhwgk_i8_oddc_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/comp/device_grouped_conv2d_fwd_xdl_ngchw_gkcyx_ngkhw_bf16_comp_2x_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/comp/device_grouped_conv2d_fwd_xdl_ngchw_gkcyx_ngkhw_bf16_comp_instance.in",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/comp/device_grouped_conv2d_fwd_xdl_ngchw_gkcyx_ngkhw_bf16_comp_part2_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/comp/device_grouped_conv2d_fwd_xdl_ngchw_gkcyx_ngkhw_f16_comp_2x_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/comp/device_grouped_conv2d_fwd_xdl_ngchw_gkcyx_ngkhw_f16_comp_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/comp/device_grouped_conv2d_fwd_xdl_ngchw_gkcyx_ngkhw_f16_comp_part2_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/comp/device_grouped_conv2d_fwd_xdl_ngchw_gkcyx_ngkhw_f32_comp_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/comp/device_grouped_conv2d_fwd_xdl_nhwgc_gkyxc_nhwgk_bf16_comp_2x_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/comp/device_grouped_conv2d_fwd_xdl_nhwgc_gkyxc_nhwgk_bf16_comp_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/comp/device_grouped_conv2d_fwd_xdl_nhwgc_gkyxc_nhwgk_bf16_comp_part2_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/comp/device_grouped_conv2d_fwd_xdl_nhwgc_gkyxc_nhwgk_f16_comp_2x_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/comp/device_grouped_conv2d_fwd_xdl_nhwgc_gkyxc_nhwgk_f16_comp_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/comp/device_grouped_conv2d_fwd_xdl_nhwgc_gkyxc_nhwgk_f16_comp_part2_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/comp/device_grouped_conv2d_fwd_xdl_nhwgc_gkyxc_nhwgk_f32_comp_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/comp/device_grouped_conv2d_fwd_xdl_nhwgc_gkyxc_nhwgk_f32_tf32_comp_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/comp/device_grouped_conv2d_fwd_xdl_nhwgc_gkyxc_nhwgk_int8_comp_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/device_grouped_conv2d_fwd_xdl_gnhwc_gkyxc_gnhwk_bf16_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/device_grouped_conv2d_fwd_xdl_gnhwc_gkyxc_gnhwk_f16_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/device_grouped_conv2d_fwd_xdl_gnhwc_gkyxc_gnhwk_f32_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/device_grouped_conv2d_fwd_xdl_ngchw_gkcyx_ngkhw_bf16_16x16_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/device_grouped_conv2d_fwd_xdl_ngchw_gkcyx_ngkhw_bf16_instance.in",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/device_grouped_conv2d_fwd_xdl_ngchw_gkcyx_ngkhw_f16_16x16_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/device_grouped_conv2d_fwd_xdl_ngchw_gkcyx_ngkhw_f16_instance.in",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/device_grouped_conv2d_fwd_xdl_ngchw_gkcyx_ngkhw_f32_16x16_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/device_grouped_conv2d_fwd_xdl_ngchw_gkcyx_ngkhw_f32_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/device_grouped_conv2d_fwd_xdl_ngchw_gkyxc_ngkhw_bf16_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/device_grouped_conv2d_fwd_xdl_ngchw_gkyxc_ngkhw_f16_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/device_grouped_conv2d_fwd_xdl_ngchw_gkyxc_ngkhw_f32_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/device_grouped_conv2d_fwd_xdl_ngchw_gkyxc_ngkhw_int8_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/device_grouped_conv2d_fwd_xdl_nhwgc_gkyxc_nhwgk_bf16_16x16_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/device_grouped_conv2d_fwd_xdl_nhwgc_gkyxc_nhwgk_bf16_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/device_grouped_conv2d_fwd_xdl_nhwgc_gkyxc_nhwgk_f16_16x16_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/device_grouped_conv2d_fwd_xdl_nhwgc_gkyxc_nhwgk_f16_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/device_grouped_conv2d_fwd_xdl_nhwgc_gkyxc_nhwgk_f32_16x16_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/device_grouped_conv2d_fwd_xdl_nhwgc_gkyxc_nhwgk_f32_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/device_grouped_conv2d_fwd_xdl_nhwgc_gkyxc_nhwgk_f32_tf32_16x16_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/device_grouped_conv2d_fwd_xdl_nhwgc_gkyxc_nhwgk_f32_tf32_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/device_grouped_conv2d_fwd_xdl_nhwgc_gkyxc_nhwgk_int8_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/large_tensor/device_grouped_conv2d_fwd_xdl_large_tensor_nhwgc_gkyxc_nhwgk_bf16_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/large_tensor/device_grouped_conv2d_fwd_xdl_large_tensor_nhwgc_gkyxc_nhwgk_f16_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/large_tensor/device_grouped_conv2d_fwd_xdl_large_tensor_nhwgc_gkyxc_nhwgk_f32_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/large_tensor/device_grouped_conv2d_fwd_xdl_large_tensor_nhwgc_gkyxc_nhwgk_f32_tf32_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/large_tensor/device_grouped_conv2d_fwd_xdl_large_tensor_nhwgc_gkyxc_nhwgk_int8_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/mem/device_grouped_conv2d_fwd_xdl_ngchw_gkcyx_ngkhw_bf16_mem_inter_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/mem/device_grouped_conv2d_fwd_xdl_ngchw_gkcyx_ngkhw_bf16_mem_intra_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/mem/device_grouped_conv2d_fwd_xdl_ngchw_gkcyx_ngkhw_f16_mem_inter_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/mem/device_grouped_conv2d_fwd_xdl_ngchw_gkcyx_ngkhw_f16_mem_intra_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/mem/device_grouped_conv2d_fwd_xdl_ngchw_gkcyx_ngkhw_f32_mem_inter_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/mem/device_grouped_conv2d_fwd_xdl_ngchw_gkcyx_ngkhw_f32_mem_intra_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/mem/device_grouped_conv2d_fwd_xdl_nhwgc_gkyxc_nhwgk_bf16_mem_inter_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/mem/device_grouped_conv2d_fwd_xdl_nhwgc_gkyxc_nhwgk_bf16_mem_intra_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/mem/device_grouped_conv2d_fwd_xdl_nhwgc_gkyxc_nhwgk_f16_mem_inter_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/mem/device_grouped_conv2d_fwd_xdl_nhwgc_gkyxc_nhwgk_f16_mem_intra_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/mem/device_grouped_conv2d_fwd_xdl_nhwgc_gkyxc_nhwgk_f32_mem_inter_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/mem/device_grouped_conv2d_fwd_xdl_nhwgc_gkyxc_nhwgk_f32_mem_intra_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/mem/device_grouped_conv2d_fwd_xdl_nhwgc_gkyxc_nhwgk_f32_tf32_mem_inter_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/mem/device_grouped_conv2d_fwd_xdl_nhwgc_gkyxc_nhwgk_f32_tf32_mem_intra_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/mem/device_grouped_conv2d_fwd_xdl_nhwgc_gkyxc_nhwgk_int8_mem_inter_instance.in",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/mem/device_grouped_conv2d_fwd_xdl_nhwgc_gkyxc_nhwgk_int8_mem_intra_instance.in",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/merged_groups/device_grouped_conv2d_fwd_xdl_merged_groups_ngchw_gkcyx_ngkhw_bf16_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/merged_groups/device_grouped_conv2d_fwd_xdl_merged_groups_ngchw_gkcyx_ngkhw_f16_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/merged_groups/device_grouped_conv2d_fwd_xdl_merged_groups_ngchw_gkcyx_ngkhw_f32_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/merged_groups/device_grouped_conv2d_fwd_xdl_merged_groups_nhwgc_gkyxc_nhwgk_bf16_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/merged_groups/device_grouped_conv2d_fwd_xdl_merged_groups_nhwgc_gkyxc_nhwgk_f16_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/merged_groups/device_grouped_conv2d_fwd_xdl_merged_groups_nhwgc_gkyxc_nhwgk_f32_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/merged_groups/device_grouped_conv2d_fwd_xdl_merged_groups_nhwgc_gkyxc_nhwgk_f32_tf32_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/merged_groups/device_grouped_conv2d_fwd_xdl_merged_groups_nhwgc_gkyxc_nhwgk_int8_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/wmma/device_grouped_conv3d_fwd_wmma_gndhwc_gkzyxc_gndhwk_f16_1x1p0_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/wmma/device_grouped_conv3d_fwd_wmma_gndhwc_gkzyxc_gndhwk_f16_1x1s1p0_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/wmma/device_grouped_conv3d_fwd_wmma_gndhwc_gkzyxc_gndhwk_f16_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/wmma/device_grouped_conv3d_fwd_wmma_gndhwc_gkzyxc_gndhwk_f16_oddc_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/wmma/device_grouped_conv3d_fwd_wmma_gndhwc_gkzyxc_gndhwk_i8_1x1p0_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/wmma/device_grouped_conv3d_fwd_wmma_gndhwc_gkzyxc_gndhwk_i8_1x1s1p0_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/wmma/device_grouped_conv3d_fwd_wmma_gndhwc_gkzyxc_gndhwk_i8_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/wmma/device_grouped_conv3d_fwd_wmma_gndhwc_gkzyxc_gndhwk_i8_oddc_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/wmma/device_grouped_conv3d_fwd_wmma_ndhwgc_gkzyxc_ndhwgk_f16_1x1p0_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/wmma/device_grouped_conv3d_fwd_wmma_ndhwgc_gkzyxc_ndhwgk_f16_1x1s1p0_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/wmma/device_grouped_conv3d_fwd_wmma_ndhwgc_gkzyxc_ndhwgk_f16_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/wmma/device_grouped_conv3d_fwd_wmma_ndhwgc_gkzyxc_ndhwgk_f16_oddc_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/wmma/device_grouped_conv3d_fwd_wmma_ndhwgc_gkzyxc_ndhwgk_i8_1x1p0_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/wmma/device_grouped_conv3d_fwd_wmma_ndhwgc_gkzyxc_ndhwgk_i8_1x1s1p0_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/wmma/device_grouped_conv3d_fwd_wmma_ndhwgc_gkzyxc_ndhwgk_i8_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/wmma/device_grouped_conv3d_fwd_wmma_ndhwgc_gkzyxc_ndhwgk_i8_oddc_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/comp/device_grouped_conv3d_fwd_xdl_ndhwgc_gkzyxc_ndhwgk_bf16_comp_instance.in",
+        "library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/comp/device_grouped_conv3d_fwd_xdl_ndhwgc_gkzyxc_ndhwgk_f16_comp_instance.in",
+        "library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/comp/device_grouped_conv3d_fwd_xdl_ndhwgc_gkzyxc_ndhwgk_f32_comp_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/comp/device_grouped_conv3d_fwd_xdl_ndhwgc_gkzyxc_ndhwgk_f32_tf32_comp_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/comp/device_grouped_conv3d_fwd_xdl_ngcdhw_gkczyx_ngkdhw_bf16_comp_2x_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/comp/device_grouped_conv3d_fwd_xdl_ngcdhw_gkczyx_ngkdhw_bf16_comp_instance.in",
+        "library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/comp/device_grouped_conv3d_fwd_xdl_ngcdhw_gkczyx_ngkdhw_bf16_comp_part2_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/comp/device_grouped_conv3d_fwd_xdl_ngcdhw_gkczyx_ngkdhw_f16_comp_2x_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/comp/device_grouped_conv3d_fwd_xdl_ngcdhw_gkczyx_ngkdhw_f16_comp_instance.in",
+        "library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/comp/device_grouped_conv3d_fwd_xdl_ngcdhw_gkczyx_ngkdhw_f16_comp_part2_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/comp/device_grouped_conv3d_fwd_xdl_ngcdhw_gkczyx_ngkdhw_f32_comp_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/device_grouped_conv3d_fwd_xdl_gndhwc_gkzyxc_gndhwk_bf16_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/device_grouped_conv3d_fwd_xdl_gndhwc_gkzyxc_gndhwk_f16_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/device_grouped_conv3d_fwd_xdl_gndhwc_gkzyxc_gndhwk_f32_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/device_grouped_conv3d_fwd_xdl_gndhwc_gkzyxc_gndhwk_int8_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/device_grouped_conv3d_fwd_xdl_ndhwgc_gkzyxc_ndhwgk_bf16_16x16_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/device_grouped_conv3d_fwd_xdl_ndhwgc_gkzyxc_ndhwgk_bf16_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/device_grouped_conv3d_fwd_xdl_ndhwgc_gkzyxc_ndhwgk_bf8_fp8_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/device_grouped_conv3d_fwd_xdl_ndhwgc_gkzyxc_ndhwgk_bf8_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/device_grouped_conv3d_fwd_xdl_ndhwgc_gkzyxc_ndhwgk_f16_16x16_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/device_grouped_conv3d_fwd_xdl_ndhwgc_gkzyxc_ndhwgk_f16_comp_fp8_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/device_grouped_conv3d_fwd_xdl_ndhwgc_gkzyxc_ndhwgk_f16_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/device_grouped_conv3d_fwd_xdl_ndhwgc_gkzyxc_ndhwgk_f32_16x16_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/device_grouped_conv3d_fwd_xdl_ndhwgc_gkzyxc_ndhwgk_f32_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/device_grouped_conv3d_fwd_xdl_ndhwgc_gkzyxc_ndhwgk_f32_tf32_16x16_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/device_grouped_conv3d_fwd_xdl_ndhwgc_gkzyxc_ndhwgk_f32_tf32_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/device_grouped_conv3d_fwd_xdl_ndhwgc_gkzyxc_ndhwgk_fp8_bf8_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/device_grouped_conv3d_fwd_xdl_ndhwgc_gkzyxc_ndhwgk_fp8_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/device_grouped_conv3d_fwd_xdl_ndhwgc_gkzyxc_ndhwgk_int8_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/device_grouped_conv3d_fwd_xdl_ngcdhw_gkczyx_ngkdhw_bf16_16x16_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/device_grouped_conv3d_fwd_xdl_ngcdhw_gkczyx_ngkdhw_bf16_instance.in",
+        "library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/device_grouped_conv3d_fwd_xdl_ngcdhw_gkczyx_ngkdhw_f16_16x16_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/device_grouped_conv3d_fwd_xdl_ngcdhw_gkczyx_ngkdhw_f16_instance.in",
+        "library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/device_grouped_conv3d_fwd_xdl_ngcdhw_gkczyx_ngkdhw_f32_16x16_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/device_grouped_conv3d_fwd_xdl_ngcdhw_gkczyx_ngkdhw_f32_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/large_tensor/device_grouped_conv3d_fwd_xdl_large_tensor_ndhwgc_gkzyxc_ndhwgk_bf16_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/large_tensor/device_grouped_conv3d_fwd_xdl_large_tensor_ndhwgc_gkzyxc_ndhwgk_f16_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/large_tensor/device_grouped_conv3d_fwd_xdl_large_tensor_ndhwgc_gkzyxc_ndhwgk_f32_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/large_tensor/device_grouped_conv3d_fwd_xdl_large_tensor_ndhwgc_gkzyxc_ndhwgk_f32_tf32_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/mem/device_grouped_conv3d_fwd_xdl_ndhwgc_gkzyxc_ndhwgk_bf16_mem_inter_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/mem/device_grouped_conv3d_fwd_xdl_ndhwgc_gkzyxc_ndhwgk_bf16_mem_intra_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/mem/device_grouped_conv3d_fwd_xdl_ndhwgc_gkzyxc_ndhwgk_f16_mem_inter_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/mem/device_grouped_conv3d_fwd_xdl_ndhwgc_gkzyxc_ndhwgk_f16_mem_intra_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/mem/device_grouped_conv3d_fwd_xdl_ndhwgc_gkzyxc_ndhwgk_f32_mem_inter_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/mem/device_grouped_conv3d_fwd_xdl_ndhwgc_gkzyxc_ndhwgk_f32_mem_intra_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/mem/device_grouped_conv3d_fwd_xdl_ndhwgc_gkzyxc_ndhwgk_f32_tf32_mem_inter_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/mem/device_grouped_conv3d_fwd_xdl_ndhwgc_gkzyxc_ndhwgk_f32_tf32_mem_intra_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/mem/device_grouped_conv3d_fwd_xdl_ngcdhw_gkczyx_ngkdhw_bf16_mem_inter_instance.in",
+        "library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/mem/device_grouped_conv3d_fwd_xdl_ngcdhw_gkczyx_ngkdhw_bf16_mem_intra_instance.in",
+        "library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/mem/device_grouped_conv3d_fwd_xdl_ngcdhw_gkczyx_ngkdhw_f16_mem_inter_instance.in",
+        "library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/mem/device_grouped_conv3d_fwd_xdl_ngcdhw_gkczyx_ngkdhw_f16_mem_intra_instance.in",
+        "library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/mem/device_grouped_conv3d_fwd_xdl_ngcdhw_gkczyx_ngkdhw_f32_mem_inter_instance.in",
+        "library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/mem/device_grouped_conv3d_fwd_xdl_ngcdhw_gkczyx_ngkdhw_f32_mem_intra_instance.in",
+        "library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/merged_groups/device_grouped_conv3d_fwd_xdl_merged_groups_ndhwgc_gkzyxc_ndhwgk_bf16_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/merged_groups/device_grouped_conv3d_fwd_xdl_merged_groups_ndhwgc_gkzyxc_ndhwgk_f16_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/merged_groups/device_grouped_conv3d_fwd_xdl_merged_groups_ndhwgc_gkzyxc_ndhwgk_f32_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/merged_groups/device_grouped_conv3d_fwd_xdl_merged_groups_ndhwgc_gkzyxc_ndhwgk_f32_tf32_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/merged_groups/device_grouped_conv3d_fwd_xdl_merged_groups_ngcdhw_gkczyx_ngkdhw_bf16_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/merged_groups/device_grouped_conv3d_fwd_xdl_merged_groups_ngcdhw_gkczyx_ngkdhw_f16_instance.cpp",
+        "library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/merged_groups/device_grouped_conv3d_fwd_xdl_merged_groups_ngcdhw_gkczyx_ngkdhw_f32_instance.cpp"
+      ],
+      "total_instantiations": 975
+    },
+    "DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle": {
+      "files": [
+        "library/include/ck/library/tensor_operation_instance/gpu/grouped_conv_fwd/device_grouped_conv_fwd_xdl_bilinear_instance.hpp",
+        "library/include/ck/library/tensor_operation_instance/gpu/grouped_conv_fwd/device_grouped_conv_fwd_xdl_binary_outelementop_instance.hpp",
+        "library/include/ck/library/tensor_operation_instance/gpu/grouped_conv_fwd/device_grouped_conv_fwd_xdl_comp_instance.hpp",
+        "library/include/ck/library/tensor_operation_instance/gpu/grouped_conv_fwd/device_grouped_conv_fwd_xdl_dynamic_op_instance.hpp",
+        "library/include/ck/library/tensor_operation_instance/gpu/grouped_conv_fwd/device_grouped_conv_fwd_xdl_instance.hpp",
+        "library/include/ck/library/tensor_operation_instance/gpu/grouped_conv_fwd/device_grouped_conv_fwd_xdl_mem_instance.hpp",
+        "library/include/ck/library/tensor_operation_instance/gpu/grouped_conv_fwd/device_grouped_conv_fwd_xdl_merged_groups_instance.hpp",
+        "library/include/ck/library/tensor_operation_instance/gpu/grouped_conv_fwd/device_grouped_conv_fwd_xdl_outelementop_instance.hpp",
+        "library/include/ck/library/tensor_operation_instance/gpu/grouped_conv_fwd/device_grouped_conv_fwd_xdl_scale_instance.hpp",
+        "library/include/ck/library/tensor_operation_instance/gpu/grouped_conv_fwd/device_grouped_conv_fwd_xdl_scaleadd_ab_instance.hpp",
+        "library/include/ck/library/tensor_operation_instance/gpu/grouped_conv_fwd/device_grouped_conv_fwd_xdl_scaleadd_scaleadd_relu_instance.hpp"
+      ],
+      "total_instantiations": 711
+    },
+    "DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3": {
+      "files": [
+        "library/include/ck/library/tensor_operation_instance/gpu/grouped_conv_fwd/device_grouped_conv_fwd_xdl_comp_instance.hpp",
+        "library/include/ck/library/tensor_operation_instance/gpu/grouped_conv_fwd/device_grouped_conv_fwd_xdl_mem_instance.hpp"
+      ],
+      "total_instantiations": 141
+    },
+    "DeviceGroupedConvFwdMultipleD_Wmma_CShuffle": {
+      "files": [
+        "library/include/ck/library/tensor_operation_instance/gpu/grouped_conv_fwd/device_grouped_conv_fwd_wmma_instance.hpp"
+      ],
+      "total_instantiations": 42
+    },
+    "DeviceGroupedConvFwdMultipleD_Xdl_CShuffle_Large_Tensor": {
+      "files": [
+        "library/include/ck/library/tensor_operation_instance/gpu/grouped_conv_fwd/device_grouped_conv_fwd_xdl_large_tensor_instance.hpp"
+      ],
+      "total_instantiations": 12
+    }
+  }
+}

forward_convolution_device_ops_report.txt

@@ -0,0 +1,290 @@
+================================================================================
+FORWARD CONVOLUTION DEVICE OPERATIONS REPORT
+================================================================================
+
+Total Unique Device Operations: 9
+
+--------------------------------------------------------------------------------
+Device Operation: DeviceConv2dFwdXdl_C_Shuffle_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K
+Total Template Instantiations: 61
+Number of Files: 1
+
+Files containing this device operation:
+  - library/src/tensor_operation_instance/gpu/conv2d_fwd/device_conv2d_fwd_xdl_c_shuffle_nhwc_kyxc_nhwk_f16_instance.cpp
+
+--------------------------------------------------------------------------------
+Device Operation: DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K
+Total Template Instantiations: 156
+Number of Files: 4
+
+Files containing this device operation:
+  - library/src/tensor_operation_instance/gpu/conv2d_fwd/device_conv2d_fwd_xdl_nhwc_kyxc_nhwk_bf16_instance.cpp
+  - library/src/tensor_operation_instance/gpu/conv2d_fwd/device_conv2d_fwd_xdl_nhwc_kyxc_nhwk_f16_instance.cpp
+  - library/src/tensor_operation_instance/gpu/conv2d_fwd/device_conv2d_fwd_xdl_nhwc_kyxc_nhwk_f32_instance.cpp
+  - library/src/tensor_operation_instance/gpu/conv2d_fwd/device_conv2d_fwd_xdl_nhwc_kyxc_nhwk_int8_instance.cpp
+
+--------------------------------------------------------------------------------
+Device Operation: DeviceConvFwd
+Total Template Instantiations: 12
+Number of Files: 6
+
+Files containing this device operation:
+  - library/include/ck/library/tensor_operation_instance/gpu/convolution_forward.hpp
+  - library/src/tensor_operation_instance/gpu/conv2d_fwd/device_conv2d_fwd_xdl_c_shuffle_nhwc_kyxc_nhwk_f16_instance.cpp
+  - library/src/tensor_operation_instance/gpu/conv2d_fwd/device_conv2d_fwd_xdl_nhwc_kyxc_nhwk_bf16_instance.cpp
+  - library/src/tensor_operation_instance/gpu/conv2d_fwd/device_conv2d_fwd_xdl_nhwc_kyxc_nhwk_f16_instance.cpp
+  - library/src/tensor_operation_instance/gpu/conv2d_fwd/device_conv2d_fwd_xdl_nhwc_kyxc_nhwk_f32_instance.cpp
+  - library/src/tensor_operation_instance/gpu/conv2d_fwd/device_conv2d_fwd_xdl_nhwc_kyxc_nhwk_int8_instance.cpp
+
+--------------------------------------------------------------------------------
+Device Operation: DeviceGroupedConvFwdDlMultipleD_NHWC_KYXC_NHWK
+Total Template Instantiations: 6
+Number of Files: 1
+
+Files containing this device operation:
+  - library/include/ck/library/tensor_operation_instance/gpu/grouped_conv_fwd/device_grouped_conv_fwd_dl_instance.hpp
+
+--------------------------------------------------------------------------------
+Device Operation: DeviceGroupedConvFwdMultipleABD
+Total Template Instantiations: 975
+Number of Files: 195
+
+Files containing this device operation:
+  - library/include/ck/library/tensor_operation_instance/gpu/grouped_conv_fwd/device_grouped_conv_fwd_xdl_bilinear_instance.hpp
+  - library/include/ck/library/tensor_operation_instance/gpu/grouped_conv_fwd/device_grouped_conv_fwd_xdl_binary_outelementop_instance.hpp
+  - library/include/ck/library/tensor_operation_instance/gpu/grouped_conv_fwd/device_grouped_conv_fwd_xdl_comp_instance.hpp
+  - library/include/ck/library/tensor_operation_instance/gpu/grouped_conv_fwd/device_grouped_conv_fwd_xdl_dynamic_op_instance.hpp
+  - library/include/ck/library/tensor_operation_instance/gpu/grouped_conv_fwd/device_grouped_conv_fwd_xdl_instance.hpp
+  - library/include/ck/library/tensor_operation_instance/gpu/grouped_conv_fwd/device_grouped_conv_fwd_xdl_mem_instance.hpp
+  - library/include/ck/library/tensor_operation_instance/gpu/grouped_conv_fwd/device_grouped_conv_fwd_xdl_merged_groups_instance.hpp
+  - library/include/ck/library/tensor_operation_instance/gpu/grouped_conv_fwd/device_grouped_conv_fwd_xdl_outelementop_instance.hpp
+  - library/include/ck/library/tensor_operation_instance/gpu/grouped_conv_fwd/device_grouped_conv_fwd_xdl_scale_instance.hpp
+  - library/include/ck/library/tensor_operation_instance/gpu/grouped_conv_fwd/device_grouped_conv_fwd_xdl_scaleadd_ab_instance.hpp
+  - library/include/ck/library/tensor_operation_instance/gpu/grouped_conv_fwd/device_grouped_conv_fwd_xdl_scaleadd_scaleadd_relu_instance.hpp
+  - library/include/ck/library/tensor_operation_instance/gpu/grouped_convolution_forward.hpp
+  - library/include/ck/library/tensor_operation_instance/gpu/grouped_convolution_forward_bias_bnorm_clamp.hpp
+  - library/include/ck/library/tensor_operation_instance/gpu/grouped_convolution_forward_bias_clamp.hpp
+  - library/include/ck/library/tensor_operation_instance/gpu/grouped_convolution_forward_bilinear.hpp
+  - library/include/ck/library/tensor_operation_instance/gpu/grouped_convolution_forward_clamp.hpp
+  - library/include/ck/library/tensor_operation_instance/gpu/grouped_convolution_forward_convinvscale.hpp
+  - library/include/ck/library/tensor_operation_instance/gpu/grouped_convolution_forward_convscale.hpp
+  - library/include/ck/library/tensor_operation_instance/gpu/grouped_convolution_forward_convscale_add.hpp
+  - library/include/ck/library/tensor_operation_instance/gpu/grouped_convolution_forward_convscale_relu.hpp
+  - library/include/ck/library/tensor_operation_instance/gpu/grouped_convolution_forward_dynamic_op.hpp
+  - library/include/ck/library/tensor_operation_instance/gpu/grouped_convolution_forward_scale.hpp
+  - library/include/ck/library/tensor_operation_instance/gpu/grouped_convolution_forward_scaleadd_ab.hpp
+  - library/include/ck/library/tensor_operation_instance/gpu/grouped_convolution_forward_scaleadd_scaleadd_relu.hpp
+  - library/include/ck/library/tensor_operation_instance/gpu/quantization/grouped_convolution_bias_forward_perchannel_quantization.hpp
+  - library/include/ck/library/tensor_operation_instance/gpu/quantization/grouped_convolution_bias_forward_perlayer_quantization.hpp
+  - library/include/ck/library/tensor_operation_instance/gpu/quantization/grouped_convolution_forward_perchannel_quantization.hpp
+  - library/include/ck/library/tensor_operation_instance/gpu/quantization/grouped_convolution_forward_perlayer_quantization.hpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv1d_fwd/xdl/device_grouped_conv1d_fwd_xdl_gnwc_gkxc_gnwk_bf16_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv1d_fwd/xdl/device_grouped_conv1d_fwd_xdl_gnwc_gkxc_gnwk_f16_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv1d_fwd/xdl/device_grouped_conv1d_fwd_xdl_gnwc_gkxc_gnwk_f32_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv1d_fwd/xdl/device_grouped_conv1d_fwd_xdl_gnwc_gkxc_gnwk_int8_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/dl/device_grouped_conv2d_fwd_dl_gnhwc_gkyxc_gnhwk_f16_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/dl/device_grouped_conv2d_fwd_dl_gnhwc_gkyxc_gnhwk_f32_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/dl/device_grouped_conv2d_fwd_dl_nhwgc_gkyxc_nhwgk_f16_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/dl/device_grouped_conv2d_fwd_dl_nhwgc_gkyxc_nhwgk_f32_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/wmma/device_grouped_conv2d_fwd_wmma_gnhwc_gkyxc_gnhwk_f16_1x1p0_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/wmma/device_grouped_conv2d_fwd_wmma_gnhwc_gkyxc_gnhwk_f16_1x1s1p0_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/wmma/device_grouped_conv2d_fwd_wmma_gnhwc_gkyxc_gnhwk_f16_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/wmma/device_grouped_conv2d_fwd_wmma_gnhwc_gkyxc_gnhwk_f16_oddc_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/wmma/device_grouped_conv2d_fwd_wmma_gnhwc_gkyxc_gnhwk_i8_1x1p0_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/wmma/device_grouped_conv2d_fwd_wmma_gnhwc_gkyxc_gnhwk_i8_1x1s1p0_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/wmma/device_grouped_conv2d_fwd_wmma_gnhwc_gkyxc_gnhwk_i8_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/wmma/device_grouped_conv2d_fwd_wmma_gnhwc_gkyxc_gnhwk_i8_oddc_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/wmma/device_grouped_conv2d_fwd_wmma_nhwgc_gkyxc_nhwgk_f16_1x1p0_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/wmma/device_grouped_conv2d_fwd_wmma_nhwgc_gkyxc_nhwgk_f16_1x1s1p0_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/wmma/device_grouped_conv2d_fwd_wmma_nhwgc_gkyxc_nhwgk_f16_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/wmma/device_grouped_conv2d_fwd_wmma_nhwgc_gkyxc_nhwgk_f16_oddc_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/wmma/device_grouped_conv2d_fwd_wmma_nhwgc_gkyxc_nhwgk_i8_1x1p0_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/wmma/device_grouped_conv2d_fwd_wmma_nhwgc_gkyxc_nhwgk_i8_1x1s1p0_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/wmma/device_grouped_conv2d_fwd_wmma_nhwgc_gkyxc_nhwgk_i8_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/wmma/device_grouped_conv2d_fwd_wmma_nhwgc_gkyxc_nhwgk_i8_oddc_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/comp/device_grouped_conv2d_fwd_xdl_ngchw_gkcyx_ngkhw_bf16_comp_2x_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/comp/device_grouped_conv2d_fwd_xdl_ngchw_gkcyx_ngkhw_bf16_comp_instance.in
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/comp/device_grouped_conv2d_fwd_xdl_ngchw_gkcyx_ngkhw_bf16_comp_part2_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/comp/device_grouped_conv2d_fwd_xdl_ngchw_gkcyx_ngkhw_f16_comp_2x_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/comp/device_grouped_conv2d_fwd_xdl_ngchw_gkcyx_ngkhw_f16_comp_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/comp/device_grouped_conv2d_fwd_xdl_ngchw_gkcyx_ngkhw_f16_comp_part2_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/comp/device_grouped_conv2d_fwd_xdl_ngchw_gkcyx_ngkhw_f32_comp_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/comp/device_grouped_conv2d_fwd_xdl_nhwgc_gkyxc_nhwgk_bf16_comp_2x_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/comp/device_grouped_conv2d_fwd_xdl_nhwgc_gkyxc_nhwgk_bf16_comp_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/comp/device_grouped_conv2d_fwd_xdl_nhwgc_gkyxc_nhwgk_bf16_comp_part2_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/comp/device_grouped_conv2d_fwd_xdl_nhwgc_gkyxc_nhwgk_f16_comp_2x_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/comp/device_grouped_conv2d_fwd_xdl_nhwgc_gkyxc_nhwgk_f16_comp_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/comp/device_grouped_conv2d_fwd_xdl_nhwgc_gkyxc_nhwgk_f16_comp_part2_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/comp/device_grouped_conv2d_fwd_xdl_nhwgc_gkyxc_nhwgk_f32_comp_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/comp/device_grouped_conv2d_fwd_xdl_nhwgc_gkyxc_nhwgk_f32_tf32_comp_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/comp/device_grouped_conv2d_fwd_xdl_nhwgc_gkyxc_nhwgk_int8_comp_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/device_grouped_conv2d_fwd_xdl_gnhwc_gkyxc_gnhwk_bf16_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/device_grouped_conv2d_fwd_xdl_gnhwc_gkyxc_gnhwk_f16_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/device_grouped_conv2d_fwd_xdl_gnhwc_gkyxc_gnhwk_f32_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/device_grouped_conv2d_fwd_xdl_ngchw_gkcyx_ngkhw_bf16_16x16_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/device_grouped_conv2d_fwd_xdl_ngchw_gkcyx_ngkhw_bf16_instance.in
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/device_grouped_conv2d_fwd_xdl_ngchw_gkcyx_ngkhw_f16_16x16_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/device_grouped_conv2d_fwd_xdl_ngchw_gkcyx_ngkhw_f16_instance.in
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/device_grouped_conv2d_fwd_xdl_ngchw_gkcyx_ngkhw_f32_16x16_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/device_grouped_conv2d_fwd_xdl_ngchw_gkcyx_ngkhw_f32_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/device_grouped_conv2d_fwd_xdl_ngchw_gkyxc_ngkhw_bf16_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/device_grouped_conv2d_fwd_xdl_ngchw_gkyxc_ngkhw_f16_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/device_grouped_conv2d_fwd_xdl_ngchw_gkyxc_ngkhw_f32_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/device_grouped_conv2d_fwd_xdl_ngchw_gkyxc_ngkhw_int8_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/device_grouped_conv2d_fwd_xdl_nhwgc_gkyxc_nhwgk_bf16_16x16_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/device_grouped_conv2d_fwd_xdl_nhwgc_gkyxc_nhwgk_bf16_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/device_grouped_conv2d_fwd_xdl_nhwgc_gkyxc_nhwgk_f16_16x16_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/device_grouped_conv2d_fwd_xdl_nhwgc_gkyxc_nhwgk_f16_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/device_grouped_conv2d_fwd_xdl_nhwgc_gkyxc_nhwgk_f32_16x16_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/device_grouped_conv2d_fwd_xdl_nhwgc_gkyxc_nhwgk_f32_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/device_grouped_conv2d_fwd_xdl_nhwgc_gkyxc_nhwgk_f32_tf32_16x16_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/device_grouped_conv2d_fwd_xdl_nhwgc_gkyxc_nhwgk_f32_tf32_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/device_grouped_conv2d_fwd_xdl_nhwgc_gkyxc_nhwgk_int8_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/large_tensor/device_grouped_conv2d_fwd_xdl_large_tensor_nhwgc_gkyxc_nhwgk_bf16_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/large_tensor/device_grouped_conv2d_fwd_xdl_large_tensor_nhwgc_gkyxc_nhwgk_f16_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/large_tensor/device_grouped_conv2d_fwd_xdl_large_tensor_nhwgc_gkyxc_nhwgk_f32_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/large_tensor/device_grouped_conv2d_fwd_xdl_large_tensor_nhwgc_gkyxc_nhwgk_f32_tf32_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/large_tensor/device_grouped_conv2d_fwd_xdl_large_tensor_nhwgc_gkyxc_nhwgk_int8_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/mem/device_grouped_conv2d_fwd_xdl_ngchw_gkcyx_ngkhw_bf16_mem_inter_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/mem/device_grouped_conv2d_fwd_xdl_ngchw_gkcyx_ngkhw_bf16_mem_intra_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/mem/device_grouped_conv2d_fwd_xdl_ngchw_gkcyx_ngkhw_f16_mem_inter_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/mem/device_grouped_conv2d_fwd_xdl_ngchw_gkcyx_ngkhw_f16_mem_intra_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/mem/device_grouped_conv2d_fwd_xdl_ngchw_gkcyx_ngkhw_f32_mem_inter_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/mem/device_grouped_conv2d_fwd_xdl_ngchw_gkcyx_ngkhw_f32_mem_intra_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/mem/device_grouped_conv2d_fwd_xdl_nhwgc_gkyxc_nhwgk_bf16_mem_inter_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/mem/device_grouped_conv2d_fwd_xdl_nhwgc_gkyxc_nhwgk_bf16_mem_intra_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/mem/device_grouped_conv2d_fwd_xdl_nhwgc_gkyxc_nhwgk_f16_mem_inter_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/mem/device_grouped_conv2d_fwd_xdl_nhwgc_gkyxc_nhwgk_f16_mem_intra_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/mem/device_grouped_conv2d_fwd_xdl_nhwgc_gkyxc_nhwgk_f32_mem_inter_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/mem/device_grouped_conv2d_fwd_xdl_nhwgc_gkyxc_nhwgk_f32_mem_intra_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/mem/device_grouped_conv2d_fwd_xdl_nhwgc_gkyxc_nhwgk_f32_tf32_mem_inter_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/mem/device_grouped_conv2d_fwd_xdl_nhwgc_gkyxc_nhwgk_f32_tf32_mem_intra_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/mem/device_grouped_conv2d_fwd_xdl_nhwgc_gkyxc_nhwgk_int8_mem_inter_instance.in
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/mem/device_grouped_conv2d_fwd_xdl_nhwgc_gkyxc_nhwgk_int8_mem_intra_instance.in
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/merged_groups/device_grouped_conv2d_fwd_xdl_merged_groups_ngchw_gkcyx_ngkhw_bf16_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/merged_groups/device_grouped_conv2d_fwd_xdl_merged_groups_ngchw_gkcyx_ngkhw_f16_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/merged_groups/device_grouped_conv2d_fwd_xdl_merged_groups_ngchw_gkcyx_ngkhw_f32_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/merged_groups/device_grouped_conv2d_fwd_xdl_merged_groups_nhwgc_gkyxc_nhwgk_bf16_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/merged_groups/device_grouped_conv2d_fwd_xdl_merged_groups_nhwgc_gkyxc_nhwgk_f16_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/merged_groups/device_grouped_conv2d_fwd_xdl_merged_groups_nhwgc_gkyxc_nhwgk_f32_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/merged_groups/device_grouped_conv2d_fwd_xdl_merged_groups_nhwgc_gkyxc_nhwgk_f32_tf32_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/xdl/merged_groups/device_grouped_conv2d_fwd_xdl_merged_groups_nhwgc_gkyxc_nhwgk_int8_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/wmma/device_grouped_conv3d_fwd_wmma_gndhwc_gkzyxc_gndhwk_f16_1x1p0_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/wmma/device_grouped_conv3d_fwd_wmma_gndhwc_gkzyxc_gndhwk_f16_1x1s1p0_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/wmma/device_grouped_conv3d_fwd_wmma_gndhwc_gkzyxc_gndhwk_f16_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/wmma/device_grouped_conv3d_fwd_wmma_gndhwc_gkzyxc_gndhwk_f16_oddc_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/wmma/device_grouped_conv3d_fwd_wmma_gndhwc_gkzyxc_gndhwk_i8_1x1p0_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/wmma/device_grouped_conv3d_fwd_wmma_gndhwc_gkzyxc_gndhwk_i8_1x1s1p0_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/wmma/device_grouped_conv3d_fwd_wmma_gndhwc_gkzyxc_gndhwk_i8_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/wmma/device_grouped_conv3d_fwd_wmma_gndhwc_gkzyxc_gndhwk_i8_oddc_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/wmma/device_grouped_conv3d_fwd_wmma_ndhwgc_gkzyxc_ndhwgk_f16_1x1p0_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/wmma/device_grouped_conv3d_fwd_wmma_ndhwgc_gkzyxc_ndhwgk_f16_1x1s1p0_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/wmma/device_grouped_conv3d_fwd_wmma_ndhwgc_gkzyxc_ndhwgk_f16_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/wmma/device_grouped_conv3d_fwd_wmma_ndhwgc_gkzyxc_ndhwgk_f16_oddc_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/wmma/device_grouped_conv3d_fwd_wmma_ndhwgc_gkzyxc_ndhwgk_i8_1x1p0_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/wmma/device_grouped_conv3d_fwd_wmma_ndhwgc_gkzyxc_ndhwgk_i8_1x1s1p0_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/wmma/device_grouped_conv3d_fwd_wmma_ndhwgc_gkzyxc_ndhwgk_i8_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/wmma/device_grouped_conv3d_fwd_wmma_ndhwgc_gkzyxc_ndhwgk_i8_oddc_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/comp/device_grouped_conv3d_fwd_xdl_ndhwgc_gkzyxc_ndhwgk_bf16_comp_instance.in
+  - library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/comp/device_grouped_conv3d_fwd_xdl_ndhwgc_gkzyxc_ndhwgk_f16_comp_instance.in
+  - library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/comp/device_grouped_conv3d_fwd_xdl_ndhwgc_gkzyxc_ndhwgk_f32_comp_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/comp/device_grouped_conv3d_fwd_xdl_ndhwgc_gkzyxc_ndhwgk_f32_tf32_comp_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/comp/device_grouped_conv3d_fwd_xdl_ngcdhw_gkczyx_ngkdhw_bf16_comp_2x_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/comp/device_grouped_conv3d_fwd_xdl_ngcdhw_gkczyx_ngkdhw_bf16_comp_instance.in
+  - library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/comp/device_grouped_conv3d_fwd_xdl_ngcdhw_gkczyx_ngkdhw_bf16_comp_part2_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/comp/device_grouped_conv3d_fwd_xdl_ngcdhw_gkczyx_ngkdhw_f16_comp_2x_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/comp/device_grouped_conv3d_fwd_xdl_ngcdhw_gkczyx_ngkdhw_f16_comp_instance.in
+  - library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/comp/device_grouped_conv3d_fwd_xdl_ngcdhw_gkczyx_ngkdhw_f16_comp_part2_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/comp/device_grouped_conv3d_fwd_xdl_ngcdhw_gkczyx_ngkdhw_f32_comp_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/device_grouped_conv3d_fwd_xdl_gndhwc_gkzyxc_gndhwk_bf16_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/device_grouped_conv3d_fwd_xdl_gndhwc_gkzyxc_gndhwk_f16_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/device_grouped_conv3d_fwd_xdl_gndhwc_gkzyxc_gndhwk_f32_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/device_grouped_conv3d_fwd_xdl_gndhwc_gkzyxc_gndhwk_int8_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/device_grouped_conv3d_fwd_xdl_ndhwgc_gkzyxc_ndhwgk_bf16_16x16_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/device_grouped_conv3d_fwd_xdl_ndhwgc_gkzyxc_ndhwgk_bf16_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/device_grouped_conv3d_fwd_xdl_ndhwgc_gkzyxc_ndhwgk_bf8_fp8_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/device_grouped_conv3d_fwd_xdl_ndhwgc_gkzyxc_ndhwgk_bf8_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/device_grouped_conv3d_fwd_xdl_ndhwgc_gkzyxc_ndhwgk_f16_16x16_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/device_grouped_conv3d_fwd_xdl_ndhwgc_gkzyxc_ndhwgk_f16_comp_fp8_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/device_grouped_conv3d_fwd_xdl_ndhwgc_gkzyxc_ndhwgk_f16_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/device_grouped_conv3d_fwd_xdl_ndhwgc_gkzyxc_ndhwgk_f32_16x16_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/device_grouped_conv3d_fwd_xdl_ndhwgc_gkzyxc_ndhwgk_f32_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/device_grouped_conv3d_fwd_xdl_ndhwgc_gkzyxc_ndhwgk_f32_tf32_16x16_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/device_grouped_conv3d_fwd_xdl_ndhwgc_gkzyxc_ndhwgk_f32_tf32_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/device_grouped_conv3d_fwd_xdl_ndhwgc_gkzyxc_ndhwgk_fp8_bf8_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/device_grouped_conv3d_fwd_xdl_ndhwgc_gkzyxc_ndhwgk_fp8_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/device_grouped_conv3d_fwd_xdl_ndhwgc_gkzyxc_ndhwgk_int8_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/device_grouped_conv3d_fwd_xdl_ngcdhw_gkczyx_ngkdhw_bf16_16x16_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/device_grouped_conv3d_fwd_xdl_ngcdhw_gkczyx_ngkdhw_bf16_instance.in
+  - library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/device_grouped_conv3d_fwd_xdl_ngcdhw_gkczyx_ngkdhw_f16_16x16_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/device_grouped_conv3d_fwd_xdl_ngcdhw_gkczyx_ngkdhw_f16_instance.in
+  - library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/device_grouped_conv3d_fwd_xdl_ngcdhw_gkczyx_ngkdhw_f32_16x16_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/device_grouped_conv3d_fwd_xdl_ngcdhw_gkczyx_ngkdhw_f32_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/large_tensor/device_grouped_conv3d_fwd_xdl_large_tensor_ndhwgc_gkzyxc_ndhwgk_bf16_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/large_tensor/device_grouped_conv3d_fwd_xdl_large_tensor_ndhwgc_gkzyxc_ndhwgk_f16_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/large_tensor/device_grouped_conv3d_fwd_xdl_large_tensor_ndhwgc_gkzyxc_ndhwgk_f32_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/large_tensor/device_grouped_conv3d_fwd_xdl_large_tensor_ndhwgc_gkzyxc_ndhwgk_f32_tf32_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/mem/device_grouped_conv3d_fwd_xdl_ndhwgc_gkzyxc_ndhwgk_bf16_mem_inter_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/mem/device_grouped_conv3d_fwd_xdl_ndhwgc_gkzyxc_ndhwgk_bf16_mem_intra_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/mem/device_grouped_conv3d_fwd_xdl_ndhwgc_gkzyxc_ndhwgk_f16_mem_inter_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/mem/device_grouped_conv3d_fwd_xdl_ndhwgc_gkzyxc_ndhwgk_f16_mem_intra_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/mem/device_grouped_conv3d_fwd_xdl_ndhwgc_gkzyxc_ndhwgk_f32_mem_inter_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/mem/device_grouped_conv3d_fwd_xdl_ndhwgc_gkzyxc_ndhwgk_f32_mem_intra_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/mem/device_grouped_conv3d_fwd_xdl_ndhwgc_gkzyxc_ndhwgk_f32_tf32_mem_inter_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/mem/device_grouped_conv3d_fwd_xdl_ndhwgc_gkzyxc_ndhwgk_f32_tf32_mem_intra_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/mem/device_grouped_conv3d_fwd_xdl_ngcdhw_gkczyx_ngkdhw_bf16_mem_inter_instance.in
+  - library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/mem/device_grouped_conv3d_fwd_xdl_ngcdhw_gkczyx_ngkdhw_bf16_mem_intra_instance.in
+  - library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/mem/device_grouped_conv3d_fwd_xdl_ngcdhw_gkczyx_ngkdhw_f16_mem_inter_instance.in
+  - library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/mem/device_grouped_conv3d_fwd_xdl_ngcdhw_gkczyx_ngkdhw_f16_mem_intra_instance.in
+  - library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/mem/device_grouped_conv3d_fwd_xdl_ngcdhw_gkczyx_ngkdhw_f32_mem_inter_instance.in
+  - library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/mem/device_grouped_conv3d_fwd_xdl_ngcdhw_gkczyx_ngkdhw_f32_mem_intra_instance.in
+  - library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/merged_groups/device_grouped_conv3d_fwd_xdl_merged_groups_ndhwgc_gkzyxc_ndhwgk_bf16_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/merged_groups/device_grouped_conv3d_fwd_xdl_merged_groups_ndhwgc_gkzyxc_ndhwgk_f16_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/merged_groups/device_grouped_conv3d_fwd_xdl_merged_groups_ndhwgc_gkzyxc_ndhwgk_f32_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/merged_groups/device_grouped_conv3d_fwd_xdl_merged_groups_ndhwgc_gkzyxc_ndhwgk_f32_tf32_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/merged_groups/device_grouped_conv3d_fwd_xdl_merged_groups_ngcdhw_gkczyx_ngkdhw_bf16_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/merged_groups/device_grouped_conv3d_fwd_xdl_merged_groups_ngcdhw_gkczyx_ngkdhw_f16_instance.cpp
+  - library/src/tensor_operation_instance/gpu/grouped_conv3d_fwd/xdl/merged_groups/device_grouped_conv3d_fwd_xdl_merged_groups_ngcdhw_gkczyx_ngkdhw_f32_instance.cpp
+
+--------------------------------------------------------------------------------
+Device Operation: DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle
+Total Template Instantiations: 711
+Number of Files: 11
+
+Files containing this device operation:
+  - library/include/ck/library/tensor_operation_instance/gpu/grouped_conv_fwd/device_grouped_conv_fwd_xdl_bilinear_instance.hpp
+  - library/include/ck/library/tensor_operation_instance/gpu/grouped_conv_fwd/device_grouped_conv_fwd_xdl_binary_outelementop_instance.hpp
+  - library/include/ck/library/tensor_operation_instance/gpu/grouped_conv_fwd/device_grouped_conv_fwd_xdl_comp_instance.hpp
+  - library/include/ck/library/tensor_operation_instance/gpu/grouped_conv_fwd/device_grouped_conv_fwd_xdl_dynamic_op_instance.hpp
+  - library/include/ck/library/tensor_operation_instance/gpu/grouped_conv_fwd/device_grouped_conv_fwd_xdl_instance.hpp
+  - library/include/ck/library/tensor_operation_instance/gpu/grouped_conv_fwd/device_grouped_conv_fwd_xdl_mem_instance.hpp
+  - library/include/ck/library/tensor_operation_instance/gpu/grouped_conv_fwd/device_grouped_conv_fwd_xdl_merged_groups_instance.hpp
+  - library/include/ck/library/tensor_operation_instance/gpu/grouped_conv_fwd/device_grouped_conv_fwd_xdl_outelementop_instance.hpp
+  - library/include/ck/library/tensor_operation_instance/gpu/grouped_conv_fwd/device_grouped_conv_fwd_xdl_scale_instance.hpp
+  - library/include/ck/library/tensor_operation_instance/gpu/grouped_conv_fwd/device_grouped_conv_fwd_xdl_scaleadd_ab_instance.hpp
+  - library/include/ck/library/tensor_operation_instance/gpu/grouped_conv_fwd/device_grouped_conv_fwd_xdl_scaleadd_scaleadd_relu_instance.hpp
+
+--------------------------------------------------------------------------------
+Device Operation: DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3
+Total Template Instantiations: 141
+Number of Files: 2
+
+Files containing this device operation:
+  - library/include/ck/library/tensor_operation_instance/gpu/grouped_conv_fwd/device_grouped_conv_fwd_xdl_comp_instance.hpp
+  - library/include/ck/library/tensor_operation_instance/gpu/grouped_conv_fwd/device_grouped_conv_fwd_xdl_mem_instance.hpp
+
+--------------------------------------------------------------------------------
+Device Operation: DeviceGroupedConvFwdMultipleD_Wmma_CShuffle
+Total Template Instantiations: 42
+Number of Files: 1
+
+Files containing this device operation:
+  - library/include/ck/library/tensor_operation_instance/gpu/grouped_conv_fwd/device_grouped_conv_fwd_wmma_instance.hpp
+
+--------------------------------------------------------------------------------
+Device Operation: DeviceGroupedConvFwdMultipleD_Xdl_CShuffle_Large_Tensor
+Total Template Instantiations: 12
+Number of Files: 1
+
+Files containing this device operation:
+  - library/include/ck/library/tensor_operation_instance/gpu/grouped_conv_fwd/device_grouped_conv_fwd_xdl_large_tensor_instance.hpp