Commit Graph

3 Commits

Author SHA1 Message Date
Changho Hwang
1bf4e8c90e connect() APIs changed to return an instance instead of a shared_ptr (#680)
The key purpose is handling all mscclpp objects' memory internally by
hiding shared pointers from user APIs.
* `Connection` class is now a wrapper of `BaseConnection` class that is
equivalent to the previous `Connection` class
* `connect()` methods now return `Connection` instead of
`std::shared_ptr<Connection>`
* Removed `connectOnSetup()` method
2025-11-15 11:40:40 -08:00
Qinghua Zhou
a38c2ee784 FP8 support for Allreduce (#646)
Add FP8 support for Allreduce on both NVIDIA and AMD platform.
Add new data type: fp8_e4m3 and fp8_e5m2

---------

Co-authored-by: Binyang Li <binyli@microsoft.com>
2025-10-27 14:51:48 -07:00
Binyang Li
70b8297c56 Revise NCCL API implementation (#617)
- Make nccl interface extensible. Customer can register their own algo
to NCCL API. User can provide customized algo selection function.
- Fallback to NCCL/RCCL if no algo is selected based on algo selection
function
- MSCCLPP interfaces now works for any scale
2025-09-26 10:08:12 -07:00