Commit Graph

24 Commits

Author SHA1 Message Date
Changho Hwang
42be3660e0 Add a new IB stack impl that doesn't use RDMA atomics (#728)
* Added configurable InfiniBand (IB) signaling mode.
`EndpointConfig::Ib::Mode` enum selects the mode (`Default`, `Host`,
`HostNoAtomic`). `Default` is equivalent to `Host` unless specified
different by envrionment `MSCCLPP_IBV_MODE`. `Host` corresponds to the
previous implementation using RDMA atomics for signaling, while
`HostNoAtomic` uses write-with-immediate instead.
* Regarding updates in Python bindings and API.
2026-02-10 01:07:53 +00:00
Binyang Li
e21513791a Address comments for PR #692 (#733)
Rename nanobind-exposed C++ types to Cpp*
Replace MSCCLPP_EXECUTION_PLAN_DIR / MSCCLPP_NATIVE_CACHE_DIR with
MSCCLPP_CACHE_DIR across C++ and Python.
2026-02-03 10:13:20 -08:00
Binyang Li
a707273701 Torch integration (#692)
Reorganize current native algorithm implementation and DSL algorithm
implementation.
Provide unified API for DSL algo and native algo and provide interface
to tune the algo
Provide interface for pytorch integration with native API and DSL

---------

Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>
Co-authored-by: Copilot <198982749+Copilot@users.noreply.github.com>
Co-authored-by: chhwang <8018170+chhwang@users.noreply.github.com>
2026-01-21 20:32:24 -08:00
Changho Hwang
8b8593ba51 Fix Python bindings and tests (#690)
Minimal fix to make things work. We need a more careful look at
preventing silent fallback of nanobind when it fails to (properly)
construct a C++ STL object with mscclpp instances.
2025-11-21 12:53:12 -08:00
Binyang Li
5acac93dbc Integrate MSCCL++ DSL to torch workload (#620)
Provides two integration ways for MSCCL++ DSL.
1. Integrate with customized communication group
2. Integrate with NCCL API

Introduce new Python APIs to make it work:
```python
mscclpp.compile # compile dsl to json based execution plan
mscclpp.ExecutionPlanRegistry.register_plan(plan) # register the compiled plan to executionPlanRegistery
mscclpp.ExecutionPlanRegistry.set_selector(selector) # set the selector, the selector will return the best execution plan based on collection, message size, world size....
```
Fix #556

---------

Co-authored-by: Caio Rocha <caiorocha@microsoft.com>
Co-authored-by: Changho Hwang <changhohwang@microsoft.com>
2025-10-29 15:39:00 -07:00
Binyang Li
b1a88d755e Pipeline fix (#645)
Co-authored-by: github-actions <github-actions@github.com>
2025-10-10 11:26:33 -07:00
Binyang Li
ddca185add Address corner case when generating version file (#641)
Address corner case for version file generation

---------

Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>
Co-authored-by: github-actions <github-actions@github.com>
2025-10-07 14:32:33 -07:00
Qinghua Zhou
16a96ea77b Support detailed version tracking that captures git repository information (#639)
#### Version Format

The package version includes the git commit hash directly in the version
string for development builds:
- **Release version**: `0.7.0`
- **Development version**: `0.7.0.dev36+g6e2360d69` (includes short
commit hash)
- **Development with uncommitted changes**:
`0.7.0.dev36+g6e2360d69.dirty`

#### Checking Version Information

After installation, you can check the version information in several
ways:

**From Python:**
```python
import mscclpp

# Access individual attributes
print(f"Version: {mscclpp.__version__}")           # Full version with commit
Version: 0.7.0.dev36+g6e2360d69

# Get as dictionary
mscclpp.version()
{'version': '0.7.0.dev46+gb0d27c58f', 'base_version': '0.7.0', 'git_commit': 'b0d27c58f'}
```

#### Version Information Details

The version tracking captures:
- **Package Version** (`mscclpp.__version__`): Full version string
including git commit (e.g., `0.7.0.dev36+g6e2360d69`)

This information is embedded during the package build process and
remains accessible even after distribution, making it easier to debug
issues and ensure reproducibility.

---------

Co-authored-by: Binyang Li <binyli@microsoft.com>
2025-09-30 09:00:33 -07:00
Changho Hwang
2eadbaf86f python doc auto generation (#605)
Add Python API references
2025-08-11 10:34:29 -07:00
Changho Hwang
ae56698d67 New semaphore constructors (#559)
More intuitive interfaces for creating semaphores and channels. Also
allows channel construction using third-party bootstrappers directly
without overriding MSCCL++ Bootstrap.
2025-07-12 00:10:46 +00:00
Changho Hwang
3565bfdf6d Renaming channels (#436)
Renamed `ProxyChannel` to `PortChannel` and `SmChannel` to
`MemoryChannel`
2025-01-24 14:25:31 -08:00
Changho Hwang
869cdba00c Manage runtime environments (#452)
* Add `Env` class that manages all runtime environments.
* Changed `NPKIT_DUMP_DIR` to `MSCCLPP_NPKIT_DUMP_DIR`.
2025-01-15 09:44:52 -08:00
Changho Hwang
f2b52c6318 Fix Python binding of exceptions (#444)
* Fixed errors to be catchable from Python code
* Skip IB tests in Python unit tests when IB ports are down
2025-01-09 11:58:23 -08:00
Changho Hwang
34945fb107 Add GpuBuffer class (#423)
* Renamed and moved mem alloc functions into the `mscclpp::detail::`
namespace (now `mscclpp::detail::gpuCalloc*<T>()`)
* Deprecated constructor-calling mem alloc functions
(`mscclpp::makeShared*<T>()` and `mscclpp::makeUnique*<T>()`)
* Added a new `mscclpp::GpuBuffer<T>()` class that should be used in
general for allocating communication buffers
* Added a new `mscclpp.utils.GpuBuffer` Python class that inherits
`cupy.ndarray` and allocates using `mscclpp::gpuMemAlloc`
* Renamed `mscclpp::memcpyCuda*<T>()` functions into
`mscclpp::gpuMemcpy*<T>()` for name consistency
* A few fixes in NVLS memory allocation
* Tackled minor compiler warnings
2025-01-07 18:40:01 -08:00
Changho Hwang
756f24c697 Revised ProxyChannel interfaces (#400)
* Renamed `ProxyChannel` -> `BaseProxyChannel` and `SimpleProxyChannel`
-> `ProxyChannel`. It makes the interface more consistent by defining
channels to be associated with a certain src/dst memory region:
`ProxyChannel` as "sema + src/dst + fifo" and `SmChannel` as "sema +
src/dst". BaseProxyChannel is not associated with any memory regions, as
"sema + fifo".
* `ProxyChannelDeviceHandle` now inherits from
`BaseProxyChannelDeviceHandle`, instead of having one as a member.
2024-12-06 10:53:34 -08:00
Binyang Li
28a57b0610 NVLS support for msccl++ executor (#375)
- Support mote datatype for multicast operation
- Add new OP MULTI_LOAD_REDUCE_STORE to support NVLS
- Modify allocSharedPhysicalCuda, which return std::shared_ptr<T>
instead of std::shared_ptr<PhysicalCudaMemory>
- Add Python support for allocSharedPhysicalCuda

Test passed for `allreduce_nvls.json`
2024-11-20 06:43:28 +00:00
Roshan Dathathri
7ed13ec4b5 Auto-tune vector sizes for NVLS allreduce6 (#338)
Also fixes bugs in MscclppAllReduce6

Below is the performance when the algorithm is fixed to
MscclppAllReduce6 on 8 H100 GPUs connected with NVLink using CUDA 12.2.

Float16:

+-------------+-----------+--------------+-------------+----------------+-------------------+------------------+----------+
| Size (fp16) | Time (us) | AlgBW (GB/s) | Correctness | NCCL Time (us)
| NCCL AlgBW (GB/s) | NCCL Correctness | Speed Up |

+-------------+-----------+--------------+-------------+----------------+-------------------+------------------+----------+
| 2.0 KiB | 11.15 | 0.18 | PASS | 13.82 | 0.15 | PASS | 1.24 |
| 4.0 KiB | 11.15 | 0.37 | PASS | 14.74 | 0.28 | PASS | 1.32 |
| 8.0 KiB | 11.14 | 0.74 | PASS | 15.17 | 0.54 | PASS | 1.36 |
| 16.0 KiB | 11.16 | 1.47 | PASS | 15.77 | 1.04 | PASS | 1.41 |
| 32.0 KiB | 11.15 | 2.94 | PASS | 17.50 | 1.87 | PASS | 1.57 |
| 64.0 KiB | 11.18 | 5.86 | PASS | 17.64 | 3.71 | PASS | 1.58 |
| 128.0 KiB | 11.16 | 11.74 | PASS | 17.83 | 7.35 | PASS | 1.60 |
| 256.0 KiB | 11.21 | 23.38 | PASS | 18.00 | 14.57 | PASS | 1.60 |
| 512.0 KiB | 11.70 | 44.81 | PASS | 18.42 | 28.46 | PASS | 1.57 |
| 1.0 MiB | 13.64 | 76.87 | PASS | 20.23 | 51.83 | PASS | 1.48 |
| 2.0 MiB | 17.29 | 121.27 | PASS | 31.60 | 66.36 | PASS | 1.83 |
| 4.0 MiB | 25.26 | 166.02 | PASS | 38.74 | 108.26 | PASS | 1.53 |
| 8.0 MiB | 40.17 | 208.83 | PASS | 62.86 | 133.45 | PASS | 1.56 |
| 16.0 MiB | 70.92 | 236.56 | PASS | 113.36 | 147.99 | PASS | 1.60 |
| 32.0 MiB | 131.38 | 255.41 | PASS | 203.21 | 165.13 | PASS | 1.55 |
| 64.0 MiB | 253.39 | 264.84 | PASS | 342.12 | 196.15 | PASS | 1.35 |
| 128.0 MiB | 496.74 | 270.20 | PASS | 670.62 | 200.14 | PASS | 1.35 |
| 256.0 MiB | 982.42 | 273.24 | PASS | 1318.36 | 203.61 | PASS | 1.34 |

+-------------+-----------+--------------+-------------+----------------+-------------------+------------------+----------+
 
 Float32:

+-------------+-----------+--------------+-------------+----------------+-------------------+------------------+----------+
| Size (fp32) | Time (us) | AlgBW (GB/s) | Correctness | NCCL Time (us)
| NCCL AlgBW (GB/s) | NCCL Correctness | Speed Up |

+-------------+-----------+--------------+-------------+----------------+-------------------+------------------+----------+
| 4.0 KiB | 11.04 | 0.37 | PASS | 14.79 | 0.28 | PASS | 1.34 |
| 8.0 KiB | 11.15 | 0.73 | PASS | 15.25 | 0.54 | PASS | 1.37 |
| 16.0 KiB | 11.12 | 1.47 | PASS | 15.87 | 1.03 | PASS | 1.43 |
| 32.0 KiB | 11.13 | 2.95 | PASS | 17.21 | 1.90 | PASS | 1.55 |
| 64.0 KiB | 11.11 | 5.90 | PASS | 17.37 | 3.77 | PASS | 1.56 |
| 128.0 KiB | 11.08 | 11.83 | PASS | 17.54 | 7.47 | PASS | 1.58 |
| 256.0 KiB | 11.15 | 23.50 | PASS | 17.71 | 14.80 | PASS | 1.59 |
| 512.0 KiB | 11.56 | 45.34 | PASS | 18.21 | 28.79 | PASS | 1.57 |
| 1.0 MiB | 13.64 | 76.90 | PASS | 19.87 | 52.77 | PASS | 1.46 |
| 2.0 MiB | 17.24 | 121.67 | PASS | 31.63 | 66.30 | PASS | 1.84 |
| 4.0 MiB | 25.19 | 166.47 | PASS | 38.63 | 108.57 | PASS | 1.53 |
| 8.0 MiB | 40.38 | 207.72 | PASS | 62.65 | 133.89 | PASS | 1.55 |
| 16.0 MiB | 70.72 | 237.23 | PASS | 114.57 | 146.44 | PASS | 1.62 |
| 32.0 MiB | 131.49 | 255.18 | PASS | 200.79 | 167.11 | PASS | 1.53 |
| 64.0 MiB | 253.98 | 264.23 | PASS | 342.58 | 195.89 | PASS | 1.35 |
| 128.0 MiB | 496.96 | 270.08 | PASS | 670.64 | 200.13 | PASS | 1.35 |
| 256.0 MiB | 982.83 | 273.12 | PASS | 1318.90 | 203.53 | PASS | 1.34 |
| 512.0 MiB | 1954.07 | 274.75 | PASS | 2609.04 | 205.77 | PASS | 1.34 |

+-------------+-----------+--------------+-------------+----------------+-------------------+------------------+----------+
2024-08-16 11:11:54 +08:00
Ziyue Yang
76328fe623 Add NPKit GPU event support (#310) 2024-06-13 13:59:50 +08:00
Binyang Li
64d837f9ab Add executor to execute schedule-plan file (#283)
Add executor to execute the JSON schedule file generated by msccl-tools

---------

Co-authored-by: Changho Hwang <changhohwang@microsoft.com>
2024-04-18 19:10:41 +00:00
Saeed Maleki
91d592dcc0 NVLS support. (#250)
Co-authored-by: Saeed Maleki <saemal@microsoft.com>
Co-authored-by: Binyang Li <binyli@microsoft.com>
Co-authored-by: Changho Hwang <changhohwang@microsoft.com>
2024-02-04 20:46:10 -08:00
Changho Hwang
060fda12e6 mscclpp-test in Python (#204)
Co-authored-by: Binyang Li <binyli@microsoft.com>
Co-authored-by: Saeed Maleki <saemal@microsoft.com>
Co-authored-by: Esha Choukse <eschouks@microsoft.com>
2023-11-16 12:45:25 +08:00
Changho Hwang
8c0f9e84d0 v0.3.0 (#171) 2023-10-11 22:35:54 +08:00
Saeed Maleki
8d1b984bed Change device handle interfaces & others (#142)
* Changed device handle interfaces
* Changed proxy service interfaces
* Move device code into separate files
* Fixed FIFO polling issues
* Add configuration arguments in several interface functions

---------

Co-authored-by: Changho Hwang <changhohwang@microsoft.com>
Co-authored-by: Binyang Li <binyli@microsoft.com>
Co-authored-by: root <root@a100-saemal0.qxveptpukjsuthqvv514inp03c.gx.internal.cloudapp.net>
2023-08-16 20:00:56 +08:00
Olli Saarikivi
4865b2017b Add Python get_include() (#141)
Introduces a mscclpp.get_include() in the Python module.
The extension module is now named _mscclpp so that we can have
Python code in the mscclpp module.
Also does some miscellaneous cleanup.
2023-07-25 10:23:16 -07:00