#3375 [Benchmark] RX 9070 XT on Windows 11, Coder IQ1_M at 64K, engine 0.1.45.3 built on the PC

本文

## Problem and result

`docs/AMD_HIP.md` asks for reports of a Windows AMD card running a model ("The ready-made zip itself has not run a model on a discrete card yet - please report", and "Reporting a Windows AMD run"). This is one: an **RX 9070 XT 16 GB (gfx1201) on Windows 11**, Ryzen 9 9950X, **31 GB RAM**, running the **Coder IQ1_M at 65,536 context** in the low-RAM mode. The engine (0.1.45.3) was built on this PC with `tools\hip\build_windows.bat` because no ready-made AMD engine was published for 0.1.45.3.

Results only, no code change. Medians and ranges from one `tools/strata_bench.py --long` suite on a fresh server:

| Test | Result |
| --- | --- |
| chat decode | **46.9 tok/s** (42.5-58.8, n=9), first token 0.64 s |
| fresh prompt 4K / 16K / 32K | 909 / 1,271 / 1,237 tok/s (n=3 each), first token 4.5 / 12.7 / 26.0 s |
| agent session, follow-up turns | decode 60.5 tok/s (25.2-69.1, n=15), cache hit 69.6 %, first token 3.4 s |

**The experimental speed projection was on** (chosen in setup); there is no run without it. Full report: `bench/results/2026-10-20-community-rx-9070-xt-qwen3-8-flash-next-coder-iq1-m/README.md`, with per-request data (`results.json`), the engine's start-up lines (`engine-log-excerpt.txt`) and the device checks (`strata-device.txt`). One entry added to the list in `docs/COMMUNITY_BENCHMARKS.md`.

## Related discussion

- The `docs/AMD_HIP.md` request above (#325 is the Windows AMD thread it names).
- #3235: the `strata_bench.py` start-up `NameError`, reproduced here (see Failed).

## Approach

Measured with the repository's own `tools/strata_bench.py`, unchanged. Because its self-start path fails (#3235), the server was started with the tool's own `start_server()` in a separate process and the suite run with `--config strata-coder-iq1_m.json --url http://127.0.0.1:8080 --long`. The tool's report was kept as written; the hardware, build, storage, low-RAM and speed-projection facts it marks "not recorded" were filled in from Windows (WMI) and the engine log. Paths in the added files are replaced by `<HOME>`, `<STRATA>`, `<DATA>`.

## Validation

**Passed:**
- `engine\strata-device.exe --list-devices` (PATH: `engine\rocm\bin` first), e88ee39d build: lists device 0 (integrated Radeon, gfx1036, "cannot run") and device 1 (RX 9070 XT, gfx1201).
- `HIP_VISIBLE_DEVICES=1 engine\strata-device.exe --selftest`: `strata-device selftest OK` on the RX 9070 XT.
- `tools/strata_bench.py --config strata-coder-iq1_m.json --url http://127.0.0.1:8080 --long`: `failures: []` and `skipped: {}` in `results.json`, suite 239 s; numbers above.
- Files checked for user names, host name, home paths, IP addresses and the server's API key: none present.

**Failed:**
- `engine\strata-device.exe --selftest` without `HIP_VISIBLE_DEVICES`, as `docs/AMD_HIP.md` gives it: exits 1 because it tests device 0, the integrated GPU (output in `strata-device.txt`).
- `python tools/strata_bench.py --long` (the self-start path `BENCH.bat` uses): `NameError: name 'port' is not defined` at `tools/strata_bench.py:990`, as in #3235.

**Not run:**
- A run without the experimental speed projection, and `STRATA_HIP_WMMA=1` (off by default on Windows).
- RAM at its rated speed: the DDR5-5600 modules ran at 4800 MT/s (EXPO off).
- More than one suite or server start; needle or other correctness checks; the ready-made 0.2.0 `rx9070` package.

## Limits

One machine, one suite run, with the Windows desktop and other apps open. The engine warned that the Windows commit limit had 4.7 GiB left at start (#1607); no crash happened. The numbers include the speed projection and the low-RAM mode on 31 GB, so they are not comparable to a run with more RAM or without the control vector.

🤖 Generated with [Claude Code](https://claude.com/claude-code)