Skip to content

[Docs] add a Windows Strix Halo (gfx1151) deployment guide - #3029

Closed
HaiYunChuang wants to merge 2 commits into
Niko1221:mainfrom
HaiYunChuang:docs/strix-halo-windows
Closed

HaiYunChuang wants to merge 2 commits into
Niko1221:mainfrom
HaiYunChuang:docs/strix-halo-windows

Conversation

@HaiYunChuang

@HaiYunChuang HaiYunChuang commented Oct 18, 2026 •

Copy link
Copy Markdown
Contributor

Problem and result

Windows users of Strix Halo (Ryzen AI Max, gfx1151) have no deployment guide: the defaults that work on Linux do not all transfer to Windows, and the setup notes are spread across issues. This PR adds a guide for a working 256K configuration and, in section 3.6, the first interleaved A/B of the gfx1151 defaults measured on Windows (engine 0.1.47).

Before: no Windows gfx1151 document. After: docs/STRIX_HALO_WINDOWS.md.

Related discussion

Issue #3032 (community reports for the same machine). No maintainer agreement on scope was requested for a docs-only change.

Approach

One new file under docs/. It records the configuration, the Windows-specific traps, the diagnostics, and the measured results. No engine, default or code change. The change touches only that file, so it applies to main as it is.

Validation

Revision tested: main at 011c9e60 (v0.2.2), engine 0.1.47 (strata-engine-0.1.47-windows-x86_64-hip-rdna3, gfx1151), Windows 11, Radeon 8060S, 128 GB. All checks below were run on that revision, after updating to it.

Passed:

  • Engine 0.1.47, Windows 11, Radeon 8060S, 128 GB: /health loaded=true, max_context 262144; /metrics engine.version=0.1.47; log session is up (engine 0.1.47); hipBLASLt tuning enabled (32 rows, gfx1151, version 100500).
  • Interleaved A/B, 3 starts per side, 1,646-token prompt, IQ3_XXS: defaults on 570.0 prefill / 56.1 decode vs STRATA_GFX1151_DEFAULTS=0 534.5 (-6.2%) / 46.5 (-17%); STRATA_REFILL_PARTS=1 561.8 (+1.0%) / 43.1; tools/hip table 594.5 / 45.8 vs the package table 576.6 (-3.0%) / 43.9 (-4.1%).
  • Vision: the CPU reader read STRATA 7391 and a red circle from a test image.
  • The file keeps UTF-8 without BOM and CRLF.

Failed:

  • none observed.

Not run:

  • Needle/correctness checks, TTFT and multi-turn agent workloads were not measured.
  • The A/B used IQ3_XXS; IQ3_S was not re-run on 0.1.47.
  • No code or engine build was tested; this is a docs-only change.

Limits

The guide reflects one machine (Ryzen AI Max+ 395 / 8060S, Windows 11, 128 GB). The 18 defaults were measured as a group; individual switches were not bisected. Absolute numbers are valid only within one interleaved run: the same configuration drifted +28% over a session.

@HaiYunChuang
HaiYunChuang marked this pull request as ready for review October 18, 2026 09:44
@mistral13

Copy link
Copy Markdown

Have you seen my community test suite reports #3032 for the Radeon 8060S (gfx1151) on a Ryzen AI Max+ 395 (Windows 11)? They’re for version 0.44.1, though.

@Niko1221

Copy link
Copy Markdown
Owner

Thanks for the pull request.

We're resetting the tracker for 0.2.0-beta. There are more than 1,300 open issues and pull requests, too many for us to review properly, so we're closing everything that isn't being worked on right now. Closed here does not mean rejected.

To reopen it takes about two minutes:

  1. Rebase on main if it no longer applies, and re-run your checks on 0.2.0-beta.
  2. Start the title with its area: [Engine], [GUI], [TUI], [Benchmark], [Docs] or [Other].
  3. Fill in the pull request template: the problem and result, and the validation (what you ran, on which revision, what you saw).
  4. Comment /reopen. A check runs and either reopens the pull request or lists what is still missing.

The rules are in CONTRIBUTING.md.

@HaiYunChuang HaiYunChuang changed the title docs: add a Windows Strix Halo (gfx1151) deployment guide [Docs] add a Windows Strix Halo (gfx1151) deployment guide Oct 22, 2026
@HaiYunChuang

Copy link
Copy Markdown
Contributor Author

/reopen

@github-actions

Copy link
Copy Markdown

Not reopened yet. Still missing:

  • Say in Validation which revision you tested: rebase on main (or 0.2.0-beta) and re-run your checks.

Edit the pull request and comment /reopen again. Details: CONTRIBUTING.md

@HaiYunChuang

Copy link
Copy Markdown
Contributor Author

/reopen

@github-actions github-actions Bot reopened this Oct 22, 2026
@github-actions

Copy link
Copy Markdown

Reopened, thank you. A maintainer will pick it up from here. If it conflicts with main, rebase it first; that speeds up review.

Niko1221 added a commit that referenced this pull request Oct 27, 2026
…n on 0.2.7 (#3029)

Structure from the guide in PR #3029 by HaiYunChuang (what the PC reaches, memory and the BIOS, the configuration, what is on by default, traps),
rewritten for the 0.2.7 install and with every number measured again: the Terminal and the gfx1151 engines built from f-027-int2 (3e468f5de), a
local signed test feed, strata launch as a standard test user on the NucBox EVO-X2 (Adrenalin 26.10.1), tools/strata_bench.py over a tunnel.
Qwen IQ3_S: 44 tok/s chat, prompts 755 / 821 / 792 tok/s at 4K / 16K / 32K. DeepSeek 2.00 bpw: 12 to 14 tok/s chat, prompts 214 to 342 tok/s.
An interleaved A/B of STRATA_SH_STREAM on Windows gfx1151 (three pairs): off is 6.9% faster in chat decode, prompts unchanged.
What launch chose and logged is quoted instead of a hand-made config. The first guide's tuning rows (engines 0.1.41 to 0.1.47) are not repeated.

Co-authored-by: HaiYunChuang <97429634+HaiYunChuang@users.noreply.github.com>
Niko1221 added a commit that referenced this pull request Oct 27, 2026
Ported PRs (authors kept) and our fixes, plus the owner's part-2 items: the restored TROUBLESHOOTING entries with a
link check, #3895 (no shared-expert fork calibration on Windows HIP with --kv-resident; a warning on a hand-set
STRATA_SH_STREAM=1), #3930 (docs), #3421 (Docker stops when models are present and MODEL_PATH is unset), #3570
(launch --show-key reads setup's config), #3311 (engine paragraph only) and #3029 (Strix Halo Windows guide,
re-measured on 0.2.7).
@Niko1221

Copy link
Copy Markdown
Owner

Thank you for the guide and the A/B table. 0.2.7 has docs/STRIX_HALO_WINDOWS.md, which follows the structure of yours and was measured again on a Ryzen AI Max+ 395 (Windows 11, Adrenalin 26.10.1) with the 0.2.7 candidate. Your finding that STRATA_SH_STREAM is slower on Windows held (decode 43.6 to 46.6 tok/s, +6.9%, same token ids), so it is off by default on Windows gfx1151 now. Fixed in 0.2.7.

Release: https://github.com/Niko1221/Strata/releases/tag/v0.2.7

@Niko1221 Niko1221 closed this Oct 27, 2026
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants