Skip to content

Pull requests: ROCm/rccl

New pull request New
26 Open 1,493 Closed
26 Open 1,493 Closed
Author
Filter by author
Loading
Label
Filter by label
Loading
Use alt + click/return to exclude labels
or ⇧ + click/return for logical OR
Projects
Filter by project
Loading
Milestones
Filter by milestone
Loading
Reviews
Filter by reviews
No reviews Review required Approved review Changes requested
Assignee
Filter by who’s assigned
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Pull requests list

Fix MSCCLPP Accuracy issue for allreduce7
#1634 opened Apr 10, 2025 by nusislam Loading…
1 task
1
Address nested designator warning issue
#1633 opened Apr 10, 2025 by mustafabar Loading…
1 task
Use internal rccl code in topo_explorer to get production tuning table
#1628 opened Apr 3, 2025 by mustafabar • Draft
1 task
Fixed unit-test env var list parsing and improved filtered test run speed
#1626 opened Apr 3, 2025 by corey-derochie-amd Loading…
1 task
@corey-derochie-amd
1
[DRAFT] Changes fp8 implementation to more closely match NCCL, and added logi…
#1619 opened Mar 27, 2025 by corey-derochie-amd • Draft
1 task
8
Enable MSCCL++ enabled UBR test for AllReduce, AllGather with TestBed::RunSimpleSweep
#1611 opened Mar 20, 2025 by isaki001 Loading…
1 task
Modified algo/proto tuning for MI300 with addition of LL128
#1609 opened Mar 19, 2025 by PedramAlizadeh Loading…
1 task
2
Add Compilation Flag for enabling/disabling clipping, and tune number of blocks for mscclpp allreduce8
#1607 opened Mar 18, 2025 by isaki001 Loading…
1 task
6
RCCL Replayer update
#1603 opened Mar 14, 2025 by AtlantaPepsi Loading…
74
RCCL deadlock discussion (was: increase NCCL_STEPS to match WARPSIZE/4)
#1600 opened Mar 13, 2025 by jglaser • Draft
1 task
25
Adding tuning for LL128 for MI300
#1590 opened Mar 5, 2025 by PedramAlizadeh Loading…
1 task
1
Remove unused variables ci:extended
#1587 opened Mar 3, 2025 by r-barnes Loading…
1 task done
1
[mscclpp] add mscclpp op mode logging
#1579 opened Feb 28, 2025 by cenzhaometa Loading…
1 task
optimize NIC permuation using bitmask DP ci:extended gfx942
#1566 opened Feb 24, 2025 by codinggosu Loading…
1 task done
10
ext-src: add initial RoCE support for MSCCLPP and Ethernet bootstrapping
#1563 opened Feb 21, 2025 by mustafabar • Draft
1 task
[EXPERIMENTAL] Add add_bf16 with rtz
#1561 opened Feb 20, 2025 by zjing14 • Draft
1 task
1
replacing rccl_float8 with hip_fp8 and address compatibility issue ci:extended
#1546 opened Feb 13, 2025 by corey-derochie-amd Loading…
1 task
@mberenjk
2
enable 224 channels
#1541 opened Feb 10, 2025 by isaki001 Loading…
1 task
7
Create internal thread to dump proxy information at given interval.
#1537 opened Feb 6, 2025 by thananon Loading…
1 task
@wenkaidu
9
Hide all warnings when compiling the mscclpp project
#1501 opened Jan 22, 2025 by corey-derochie-amd Loading…
1 task
@corey-derochie-amd
Adds API for non-caching-store and load
#1476 opened Jan 2, 2025 by ranapratap55 Loading…
9
Fix synchronization in allreduce8Read kernel
#1457 opened Dec 10, 2024 by dsidler Loading…
2
[FORMAT] Add .clang-format for C++ code
#1404 opened Nov 3, 2024 by nileshnegi Loading…
1 task
1
MSCCL Multithreaded regression alternative state management
#1352 opened Sep 26, 2024 by corey-derochie-amd Loading…
1 task
@corey-derochie-amd
changing balanced-tree to tree
#1351 opened Sep 26, 2024 by PedramAlizadeh Loading…
1 task
1
ProTip! Exclude everything labeled bug with -label:bug.