From 1c700083b2eb22105d01cef809db3404854d894d Mon Sep 17 00:00:00 2001 From: corey-derochie-amd <161367113+corey-derochie-amd@users.noreply.github.com> Date: Wed, 23 Oct 2024 13:49:40 -0600 Subject: [PATCH] Update CHANGELOG to match release branches 6.2 and 6.3 (#1391) * [CHANGELOG] Add Known issues for ROCm 6.2.1 Signed-off-by: nileshnegi * Updated 6.2.1 known issues to match the content in develop. * Updated CHANGELOG for ROCm 6.3 release. (#1380) * Updated CHANGELOG for ROCm 6.3 release. * Update CHANGELOG to new format. Co-authored-by: Jeffrey Novotny --------- Co-authored-by: Jeffrey Novotny --------- Signed-off-by: nileshnegi Co-authored-by: nileshnegi Co-authored-by: Jeffrey Novotny [ROCm/rccl commit: 6ed513e1b992d2b81819b3e8cb88a621e2dedbbf] --- projects/rccl/CHANGELOG.md | 43 +++++++++++++++++++++++++++++--------- 1 file changed, 33 insertions(+), 10 deletions(-) diff --git a/projects/rccl/CHANGELOG.md b/projects/rccl/CHANGELOG.md index d753056f0d..af9b7194c9 100644 --- a/projects/rccl/CHANGELOG.md +++ b/projects/rccl/CHANGELOG.md @@ -2,19 +2,42 @@ Full documentation for RCCL is available at [https://rccl.readthedocs.io](https://rccl.readthedocs.io) +## RCCL 2.21.5 for ROCm 6.3.0 + +### Added + +* MSCCL++ integration for specific contexts +* Performance collection to rccl_replayer +* Tuner Plugin example for MI300 +* Tuning table for large number of nodes +* Support for amdclang++ +* New Rome model + +### Changed + +* Compatibility with NCCL 2.21.5 +* Increased channel count for MI300X multi-node +* Enabled MSCCL for single-process multi-threaded contexts +* Enabled gfx12 +* Enabled CPX mode for MI300X +* Enabled tracing with rocprof +* Improved version reporting +* Enabled GDRDMA for Linux kernel 6.4.0+ + +### Resolved issues + +* Fixed model matching with PXN enable + ## RCCL 2.20.5 for ROCm 6.2.1 - +### Fixed +- GDR support flag now set with DMABUF ### Known issues +- On systems running Linux kernel 6.8.0, such as Ubuntu 24.04, Direct Memory Access (DMA) transfers between the GPU and NIC are disabled and impacts multi-node RCCL performance. + - This issue was reproduced with RCCL 2.20.5 (ROCm 6.2.0 and 6.2.1) on systems with Broadcom Thor-2 NICs and affects other systems with RoCE networks using Linux 6.8.0 or newer. + - Older RCCL versions are also impacted. + - This issue will be addressed in a future ROCm release. -On systems running Linux kernel 6.8.0, such as Ubuntu 24.04, Direct Memory Access (DMA) transfers between the GPU and NIC are disabled and impacts multi-node RCCL performance. - -This issue was reproduced with RCCL 2.20.5 (ROCm 6.2.0 and 6.2.1) on systems with Broadcom Thor-2 NICs and affects other systems with RoCE networks using Linux 6.8.0 or newer. - -Older RCCL versions are also impacted. - -This issue will be addressed in a future ROCm release. - -## Unreleased - RCCL 2.20.5 for ROCm 6.2.0 +## RCCL 2.20.5 for ROCm 6.2.0 ### Changed - Compatibility with NCCL 2.20.5 - Compatibility with NCCL 2.19.4