ROCm/composable_kernel
1
0
Fork 0
mirror of https://github.com/ROCm/composable_kernel.git synced 2026-05-01 20:21:23 +00:00
Code Issues Packages Projects Releases Wiki Activity

Ck tile engine preshuffle (#2919)

Browse Source 
* Partial Progress : Preshuffle working code for datatype

* Partial Progress : Preshuffle Cleanup

* Working code for default config with min max step

* Partial Progress : PermuteN implemented in validation

* Partial Progress : PermuteN changes in Preshuffle

* CK Tile Engine Preshuffle Complete

* CK TILE ENGINE : Preshuffle Layout validation

* CK Tile Engine Preshuffle Validation

* Preshuffle Validation check

* CK Tile Engine Preshuffle : Fixing Validation Cases

* Addressing PR review Comments

* Changes in config

* Addressing Review Comments

* Adding additional architecture in Jenkins

* Partial Progress : Selective Datatype and layouts

* Limited datatypes and layouts

* Addressing CI errors

* Datatype updates

* Datatype updates

* Datatype changes to Preshuffle

* Addressing Review Comments

* Addressing Review Comments

* Datatype changes

* Changes to Cmake

* Update on Jenkins

* Formatting with precommit

* Ruff Formatting
This commit is contained in:
Thrupti Raj Lakshmana Gowda
2025-10-27 09:15:34 -05:00
committed by GitHub
parent 6d709dac41
commit 8b185e872e
12 changed files with 393 additions and 209 deletions
Download Patch File Download Diff File Expand all files Collapse all files

81
tile_engine/ops/gemm_preshuffle/configs/default_config.json
View File

@@ -1,62 +1,72 @@
{
"tile_config": {
"tile_m": {
"values": [
128
]
"max": 256,
"min": 64,
"step": 64
},
"tile_n": {
"values": [
128
]
"max": 256,
"min": 64,
"step": 64
},
"tile_k": {
"values": [
128
]
"max": 256,
"min": 64,
"step": 64
},
"warp_m": {
"values": [
1
]
"values": [
4,
2,
1
]
},
"warp_n": {
"values": [
4
]
"values": [
4,
2,
1
]
},
"warp_k": {
"values": [
1
]
"values": [
1
]
},
"warp_tile_m": {
"values": [
16
]
"values": [
4,
16,
32
]
},
"warp_tile_n": {
"values": [
16
]
"values": [
16,
32,
64
]
},
"warp_tile_k": {
"values": [
16,32
]
"values": [
8,
16,
32,
64,
128
]
}
},
"trait_config": {
"pipeline": {
"values": [
"preshufflev1",
"preshufflev2"
]
},
"scheduler": {
"values": [
"interwave",
"intrawave"
"default"
]
},
"epilogue": {
@@ -81,11 +91,12 @@
]
},
"persistent": {
"values": [
true,
false
]
"values": [
true,
false
]
}
},
"k_block_per_cu": 2
"k_block_per_cu": 1,
"permute_n": true
}