Skip to content
Merged
Changes from 1 commit
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
25 changes: 14 additions & 11 deletions projects/rocshmem/CHANGELOG.md
Original file line number Diff line number Diff line change
Expand Up @@ -10,20 +10,23 @@
* `rocshmem_buffer_register/unregister` for GDA backend
* `rocshmem_reduce_on_stream`
* `rocshmem_team_split_2D`
* Tile-granular RMA operations for the IPC backend
* Host-initiated RMA operations in the IPC backend for the non-MPI
bootstrapping path
* Team creation using non-contiguous parent teams in the IPC backend
* Python bindings of memory-management APIs
* Python bindings coverage for team APIs
* Performance optimizations:
* Separate put/get memcpy primitives to apply correct cache coherence semantics and fences
* O(1) IPC availability check using pattern detection
* Use constmem for backend variables and provider muxing
* Added tile-granular RMA operations for the IPC backend
Comment thread
yugang-amd marked this conversation as resolved.
* Added host-initiated RMA operations in the IPC backend for the non-MPI
bootstrapping path
* Added team creation using non-contiguous parent teams in the IPC backend
* Added Python bindings of memory-management APIs
* Added Python bindings coverage for team APIs
* Added support for GPU initiated operations using the SDMA engines
* Added ASAN build support
* Change default ROCSHMEM_DEBUG_LEVEL from WARN to ERROR
* Added O(1) IPC availability check using pattern detection
Comment thread
abouteiller marked this conversation as resolved.
Outdated

### Changed
* Changed default `ROCSHMEM_DEBUG_LEVEL` from `WARN` to `ERROR`
* Performance optimizations:
* Separated put/get memcpy primitives to apply correct cache coherence semantics and fences
* Use constmem for backend variables and provider muxing


Comment thread
Copilot marked this conversation as resolved.
Outdated
## rocSHMEM 3.4.0 for ROCm 7.13
### Added
* Added new APIs:
Expand Down
Loading