[Feature]: follow-up to PyTorch allocation provider to accept already rendezvoused tensors
Maintainers usually reply within 3 days
Nobody has claimed this yet.
Assessment
- Difficulty
- 5/5
- Estimated time
- Over a week
- Newbie friendliness
- 15/100
- Issue type
- Feature
- Clarity
- Needs clarification
- Activity status
- Active
- Domain
- distributed-systems
Research direction
The payload names no source files or tests. It refers to the earlier PyTorch allocation provider in #551, torch.distributed._symmetric_memory, and the single-map iris.copy limitation, so start by reading #551 and the provider it added. Done would need an agreed API for accepting pre-rendezvoused tensors, which the issue does not yet define, so settle that design with maintainers first.
Written by the indexing model from the issue text.
Description
Suggestion Description
#551 provides capability to run Iris device kernels with memory allocated via PyTorch, with no copy, no second heap, and no kernel changes.
and we agreed on extending it to accept tensors frameworks have already rendezvoused , because users might want to manage the memory.
Overall Motivation behind #551 and this feature request is as follows:
- Some of the inference stack already uses torch.distributed.symmetric_memory. TokenSpeed's amd_rsag* path, the TritonAllReduceBackend, and the torch.ops.symm_mem.* collectives that vLLM and SGLang-style engines may call all allocate through it. Until now, using an Iris kernel meant moving those buffers into the Iris heap. This provider lets the two coexist: a framework keeps its torch allocations, and individual Iris kernels can run on them.
- Frameworks that already own their buffers can't use Iris device kernels until the follow-up lands.
- The single-map
iris.copylimitation also means it does not yet make every Iris operation work across independently allocated tensors
Operating System
No response
GPU
No response
ROCm Component
No response
- Dominant language
- Python
- Stars
- 202
- Forks
- 52
- Avg merge
- 5d 2h
- Merged PRs (30d)
- 12
Getting set up
Starts the project's dev container in your browser, under your own GitHub account.
- Ships a Dockerfile or Docker Compose file
- No pull request template
- Read the contributing guide
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
More from ROCm/iris
-
benchmarks bug examples iris test
Difficulty 2/5 1-3 hours Newbie friendliness 75/100
Maintainers usually reply within 3 days
-
bug examples invalid iris
Difficulty 2/5 1-3 hours Newbie friendliness 70/100
Maintainers usually reply within 3 days
-
benchmarks gluon iris
Difficulty 5/5 Over a week Newbie friendliness 30/100
Maintainers usually reply within 3 days
-
bug iris test
Difficulty 4/5 3-5 days Newbie friendliness 40/100
Maintainers usually reply within 3 days
-
[Feature]: Formalize allocator-agnostic symmetric tensor address translationMay be free again @artulab claimed this 49 days ago, and no pull request is open. Openiris
ROCm/iris#546 · 2 comments · 1 assignee ·
Maintainers usually reply within 3 days
Similar issues
-
[Bug]: LangGraph Store sync put bypasses namespace validationPossibly taken @ktz03 claimed this today. Open
Difficulty 2/5 1-3 hours Newbie friendliness 72/100
volcengine/OpenViking#6050 ·
Maintainers usually reply within 1 day
-
[Bug]: Streamed /v1/chat/completions sends unredacted agent error text (keys) in the final chunkPossibly taken @rodricksz4h5 claimed this today. Openarea/streaming comp/gateway P2 type/security
Difficulty 2/5 1-3 hours Newbie friendliness 78/100
NousResearch/hermes-agent#138245 ·
Maintainers usually reply within 1 day
-
[oblt-aw][autodoc] Update user onboarding guide for mandatory trigger CODEOWNERS stepPossibly taken @github-actions claimed this today. Open
Difficulty 2/5 1-3 hours Newbie friendliness 66/100
Maintainers usually reply within 1 day
-
correction metadata
Difficulty 2/5 Under an hour Newbie friendliness 68/100
acl-org/acl-anthology#10378 · 1 comment ·
Maintainers usually reply within 1 day
-
cloudtrail_multi_region_enabled_logging_management_events fails when the advanced event selector has a custom namePossibly taken @12YYCCYY16 claimed this today. Open
Difficulty 2/5 1-3 hours Newbie friendliness 78/100
prowler-cloud/prowler#13033 ·
Maintainers usually reply within 1 day