fix(ck-tile): prefer amd-smi over rocm-smi in remaining TE GEMM benchmarks by ozturkosu · Pull Request #10204 · ROCm/rocm-libraries

ozturkosu · 2026-07-30T21:26:37Z

Motivation

PR #9731 centralized GPU SMI access behind tile_engine/ops/common/smi_utils.py, which prefers amd-smi and transparently falls back to rocm-smi, and migrated gemm_full_benchmark.py off direct rocm-smi calls. Three Tile Engine benchmark scripts were left calling rocm-smi --showid directly (with an amd-smi list fallback), so the migration is incomplete. Although wrapped in try/except today, these direct calls will break once rocm-smi is removed from future ROCm releases.

This completes the migration for the remaining scripts, as requested by the tech lead and tracked in JIRA ROCM-28734.

Technical Details

Migrated the identical hand-rolled detect_devices() (env → rocm-smi --showid → amd-smi list → hardcoded ["0"]) to a single smi_utils.detect_gpu_ids() call, mirroring the already-landed gemm_full_benchmark.py:

tile_engine/ops/gemm/block_scale_gemm/gemm_bquant/gemm_bquant_full_benchmark.py
tile_engine/ops/gemm/gemm_multi_d_full_benchmark.py
tile_engine/ops/gemm/streamk_gemm_full_benchmark.py

Each script adds common/ to sys.path, imports detect_gpu_ids, and drops the now-unused import re. Net -67 lines.

Test Plan / Result

Validated on an MI350 (gfx950) host, exercising the exact sys.path arithmetic + smi_utils import + detect_gpu_ids() for both the shallow (gemm/) and deep (block_scale_gemm/gemm_bquant/) script locations:

Default (amd-smi first): detect_devices() returns the live GPU ids; count_gpus()/check_gpu_available() consistent.
CK_SMI_TOOL=rocm-smi (forced fallback): identical result.

Both python3 -m py_compile cleanly.

Submission Checklist

Look over the contributing guidelines at https://github.com/ROCm/ROCm/blob/develop/CONTRIBUTING.md#pull-requests.

https://amd-hub.atlassian.net/jira/software/c/projects/AICK/boards/4399?selectedIssue=ROCM-28734

JIRA ID : ROCM-28734

…marks The bquant, multi-D, and stream-K full-benchmark scripts still called `rocm-smi --showid` directly (with an `amd-smi list` fallback) for GPU id detection, duplicating the detection logic that PR #9731 centralized in `tile_engine/ops/common/smi_utils.py`. This completes the SMI migration started in #9731 so every call site prefers `amd-smi` and transparently falls back to `rocm-smi`. Each `detect_devices()` now delegates to `smi_utils.detect_gpu_ids()`, matching the already-migrated `gemm_full_benchmark.py`. Added the `common/` dir to `sys.path` and dropped the now-unused `import re`. Validated on an MI350 (gfx950) host: both the default (amd-smi first) and `CK_SMI_TOOL=rocm-smi` fallback paths return identical GPU ids for both the shallow (gemm/) and deep (block_scale_gemm/gemm_bquant/) script locations. JIRA ID : ROCM-28734

therock-pr-bot · 2026-07-30T21:30:51Z

✅ All Checks Passed — Ready for Review

Check	Status	Details
📝 PR Description	✅ Pass	—
⛔ Forbidden Files	✅ Pass	—
🧪 Unit Test	⚠️ Warning	Error: Source/code files changed without an accompanying unit test. Expected: add at least one test file named like `test_<name>.py` / `test_<name>.cpp` (or `<name>_test.`). Current:* code file(s) changed: `projects/composablekernel/tile_engine/ops/gemm/block_scale_gemm/gemm_bquant/gemm_bquant_full_benchmark.py`, `projects/composablekernel/tile_engine/ops/gemm/gemm_multi_d_full_benchmark.py`, `projects/composablekernel/tile_engine/ops/gemm/streamk_gemm_full_benchmark.py`; no test file found
🔎 pre-commit	✅ Pass	—
🚫 Draft PR	🔜 To Be Enabled	—
🚩 Feature Flag	🔜 To Be Enabled	—
📊 Code Coverage	🔜 To Be Enabled	—
🤖 therock-pr-bot	✅ Pass	—

🎉 All checks passed! This PR is ready for review.

📖 Need help? See the Policy FAQ for details on every check and how to fix failures.

🙋 Wish to Override Policy?

therock-pr-bot · 2026-07-30T21:30:53Z

🎉 All checks passed! This PR is ready for review.

fix(ck-tile): prefer amd-smi over rocm-smi in remaining TE GEMM benchmarks (#10204) MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit ## Motivation PR #9731 centralized GPU SMI access behind `tile_engine/ops/common/smi_utils.py`, which prefers `amd-smi` and transparently falls back to `rocm-smi`, and migrated `gemm_full_benchmark.py` off direct `rocm-smi` calls. Three Tile Engine benchmark scripts were left calling `rocm-smi --showid` directly (with an `amd-smi list` fallback), so the migration is incomplete. Although wrapped in try/except today, these direct calls will break once `rocm-smi` is removed from future ROCm releases. This completes the migration for the remaining scripts, as requested by the tech lead and tracked in JIRA **ROCM-28734**. ## Technical Details Migrated the identical hand-rolled `detect_devices()` (env → `rocm-smi --showid` → `amd-smi list` → hardcoded `["0"]`) to a single `smi_utils.detect_gpu_ids()` call, mirroring the already-landed `gemm_full_benchmark.py`: - `tile_engine/ops/gemm/block_scale_gemm/gemm_bquant/gemm_bquant_full_benchmark.py` - `tile_engine/ops/gemm/gemm_multi_d_full_benchmark.py` - `tile_engine/ops/gemm/streamk_gemm_full_benchmark.py` Each script adds `common/` to `sys.path`, imports `detect_gpu_ids`, and drops the now-unused `import re`. Net -67 lines. ## Test Plan / Result Validated on an MI350 (gfx950) host, exercising the exact `sys.path` arithmetic + `smi_utils` import + `detect_gpu_ids()` for both the shallow (`gemm/`) and deep (`block_scale_gemm/gemm_bquant/`) script locations: - Default (amd-smi first): `detect_devices()` returns the live GPU ids; `count_gpus()`/`check_gpu_available()` consistent. - `CK_SMI_TOOL=rocm-smi` (forced fallback): identical result. Both `python3 -m py_compile` cleanly. ## Submission Checklist - [x] Look over the contributing guidelines at https://github.com/ROCm/ROCm/blob/develop/CONTRIBUTING.md#pull-requests. https://amd-hub.atlassian.net/jira/software/c/projects/AICK/boards/4399?selectedIssue=ROCM-28734 JIRA ID : ROCM-28734 Co-authored-by: Muhammed Emin Ozturk <3836908+ozturkosu@users.noreply.github.com>

## Summary Extends the amd-smi-first migration (#9731, #10204) from the Tile Engine to the dispatcher's GPU-architecture detection. - Adds `detect_gpu_arch()` to the shared `tile_engine/ops/common/smi_utils.py` wrapper: amd-smi `static` (`TARGET_GRAPHICS_VERSION`) first, rocm-smi `--showproductname` (`GFX Version`) fallback — same amd-smi-first policy as the rest of the wrapper. - `dispatcher_common.detect_gpu_arch()`: prefers amd-smi via the wrapper, then rocminfo, then the caller-supplied default. Also removes a **duplicate** definition of this function. - `gemm_utils._get_arch()`: tries amd-smi via the wrapper before rocminfo, preserving the raise-on-unresolved (no silent default) behavior. ## Motivation Per the directive behind #9731, device-property access should prefer `amd-smi` (rocm-smi is being phased out). The Tile Engine was already migrated (#9731, #10204); the dispatcher still detected the GPU arch by shelling out to `rocminfo` directly. Tracked in JIRA **ROCM-28734**. ## Design note The dispatcher reaches the sibling `tile_engine/ops/common/smi_utils.py` through a guarded import and falls back to `rocminfo` if the wrapper is unavailable, so behavior is unchanged on hosts without amd-smi. (Open to moving the shared wrapper to a neutral location if reviewers prefer avoiding the dispatcher → tile_engine reach.) ## Test plan (MI350X / gfx950) - [x] New mock-based unit tests for `detect_gpu_arch()` (amd-smi preference, rocm-smi fallback, default) + a live check — 4/4 pass. - [x] `dispatcher_common.detect_gpu_arch()` and `gemm_utils._get_arch()` both return `gfx950` via amd-smi on the live host. - [x] `test_gemm_utils` 21/21 still pass (no arch-detection regression). --------- Co-authored-by: Muhammed Emin Ozturk <3836908+ozturkosu@users.noreply.github.com>

fix(ck-tile): prefer amd-smi over rocm-smi in remaining TE GEMM benchmarks ## Motivation PR #9731 centralized GPU SMI access behind `tile_engine/ops/common/smi_utils.py`, which prefers `amd-smi` and transparently falls back to `rocm-smi`, and migrated `gemm_full_benchmark.py` off direct `rocm-smi` calls. Three Tile Engine benchmark scripts were left calling `rocm-smi --showid` directly (with an `amd-smi list` fallback), so the migration is incomplete. Although wrapped in try/except today, these direct calls will break once `rocm-smi` is removed from future ROCm releases. This completes the migration for the remaining scripts, as requested by the tech lead and tracked in JIRA **ROCM-28734**. ## Technical Details Migrated the identical hand-rolled `detect_devices()` (env → `rocm-smi --showid` → `amd-smi list` → hardcoded `["0"]`) to a single `smi_utils.detect_gpu_ids()` call, mirroring the already-landed `gemm_full_benchmark.py`: - `tile_engine/ops/gemm/block_scale_gemm/gemm_bquant/gemm_bquant_full_benchmark.py` - `tile_engine/ops/gemm/gemm_multi_d_full_benchmark.py` - `tile_engine/ops/gemm/streamk_gemm_full_benchmark.py` Each script adds `common/` to `sys.path`, imports `detect_gpu_ids`, and drops the now-unused `import re`. Net -67 lines. ## Test Plan / Result Validated on an MI350 (gfx950) host, exercising the exact `sys.path` arithmetic + `smi_utils` import + `detect_gpu_ids()` for both the shallow (`gemm/`) and deep (`block_scale_gemm/gemm_bquant/`) script locations: - Default (amd-smi first): `detect_devices()` returns the live GPU ids; `count_gpus()`/`check_gpu_available()` consistent. - `CK_SMI_TOOL=rocm-smi` (forced fallback): identical result. Both `python3 -m py_compile` cleanly. ## Submission Checklist - [x] Look over the contributing guidelines at https://github.com/ROCm/ROCm/blob/develop/CONTRIBUTING.md#pull-requests. https://amd-hub.atlassian.net/jira/software/c/projects/AICK/boards/4399?selectedIssue=ROCM-28734 JIRA ID : ROCM-28734 Co-authored-by: Muhammed Emin Ozturk <3836908+ozturkosu@users.noreply.github.com>

github-actions Bot added the project: composablekernel label Jul 30, 2026

assistant-librarian Bot added the organization: ROCm label Jul 30, 2026

ozturkosu self-assigned this Jul 30, 2026

ozturkosu marked this pull request as ready for review July 30, 2026 21:35

ozturkosu requested a review from a team as a code owner July 30, 2026 21:35

ozturkosu requested a review from geyyer July 30, 2026 21:49

geyyer approved these changes Aug 4, 2026

View reviewed changes

geyyer merged commit 9116b6d into develop Aug 4, 2026
32 checks passed

geyyer deleted the users/muozturk/ck/smi-wrapper-benchmarks branch August 4, 2026 15:35

ozturkosu mentioned this pull request Aug 5, 2026

fix(ck-tile): prefer amd-smi for dispatcher GPU-arch detection #10445

Merged

3 tasks done

Navigation Menu

Repository navigation

fix(ck-tile): prefer amd-smi over rocm-smi in remaining TE GEMM benchmarks - #10204

fix(ck-tile): prefer amd-smi over rocm-smi in remaining TE GEMM benchmarks#10204
geyyer merged 1 commit into
developfrom
users/muozturk/ck/smi-wrapper-benchmarks

ozturkosu commented Jul 30, 2026 •

edited

Loading

Uh oh!

therock-pr-bot Bot commented Jul 30, 2026 •

edited

Loading

Uh oh!

therock-pr-bot Bot commented Jul 30, 2026

Uh oh!

Uh oh!

Reviewers

Uh oh!

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

Conversation

ozturkosu commented Jul 30, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Motivation

Technical Details

Test Plan / Result

Submission Checklist

Uh oh!

therock-pr-bot Bot commented Jul 30, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

✅ All Checks Passed — Ready for Review

Uh oh!

therock-pr-bot Bot commented Jul 30, 2026

Uh oh!

Uh oh!

Reviewers

Uh oh!

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

ozturkosu commented Jul 30, 2026 •

edited

Loading

therock-pr-bot Bot commented Jul 30, 2026 •

edited

Loading