Native vLLM + ROCm 7.15 Runtime for RX 6000 (RDNA2) on Windows 11 — 26 TFLOPS FP16, 62 tok/s, One-Click Install, No WSL2 [RX 6750 XT gfx1031 Verified]

Esseba-dev — Tue, 18 Aug 2026 10:02:38 +0000

I built a native vLLM + ROCm 7.15 runtime for AMD RX 6000 Series on Windows 11 — now with a one-click installer.

AMD lists RX 6750 XT / 6700 XT / 6600 XT as "Runtime only" on Windows with HIP SDK excluded. I built rocBLAS binaries for gfx1031 via ROCm/TheRock to close that gap.

This runs native Windows HIP and ROCm directly, no WSL2 wrapper, no compiler needed.

Tested on AMD Radeon RX 6750 XT 12GB (gfx1031) — Windows 11 Native — August 2026

Verification — Real terminal logs

Environment:

torch 2.12.0+rocm7.15.0 | cuda_avail True | dev AMD Radeon RX 6750 XT
vLLM 0.19.1 | plugin vllm_windows_rocm activated | TRITON_ATTN | enforce_eager=True
rocBLAS Benchmark — 26 TFLOPS FP16:

Device ID 0 : AMD Radeon RX 6750 XT gfx1031 with 12.9 GB memory
rocBLAS version: 5.7.0.67811f1ee52
transA,transB,M,N,K,alpha,lda,beta,ldb,ldc,cold_iters,hot_iters,rocblas-Gflops,us
N,N,4096,4096,4096,1,4096,0,4096,4096, 2, 10, 25977.3, 5290.73
=> 25.97 TFLOPS in 5.29ms
vLLM Inference (Qwen3.5-4B, web chat UI):

Output: 59-62 tok/s | Init ~1s | hidden thinking (spinner) then answer
Model auto-detected via /v1/models

Full logs in benchmarks/ and screenshots in assets/ on GitHub.

How it works

TheRock builds clr (HIP) and rocBLAS with Tensile kernels for gfx1031
HSA_OVERRIDE_GFX_VERSION=10.3.0 forces HIP to recognize RX 6750 XT
PyTorch 2.12.0+rocm7.15 links against TheRock runtime => torch.cuda.is_available() True
vLLM plugin vllm_windows_rocm bypasses vllm._C dependency and registers WinRocmAwqGemvKernel with TRITON_ATTN
vLLM engine loads with enforce_eager=True and runs native

Quick Start — One click

Prerequisites: Windows 11 23H2+, RX 6600-6750 XT (gfx1030/gfx1031/gfx1032), Adrenalin 24.x+, ~8GB free disk

Download the repo (or git clone) and run INSTALL.bat as Administrator
The installer downloads the release archives automatically (multi-part join included, no manual steps) and sets up C:\TheRock\ + Python 3.11 + venv
Run CHAT.bat — the browser opens the chat UI. Expected: 59-62 tok/s

VERIFY.bat runs a 3-in-1 check: GPU detection + rocBLAS benchmark + vLLM smoke test.

What's new in v2.0

One-click installer: INSTALL.bat auto-downloads and joins split archives (no browser upload/download hassle)
Chat web UI (OpenAI-compatible API): hidden thinking with spinner, then the answer streams — like NVIDIA ChatRTX
Fat-binary gemv kernels for all RDNA2 (gfx1030/1031/1032) — no per-card rebuild
ROCm 7.15 native, vLLM 0.19.1, torch 2.12.0+rocm7.15
Faster: 59-62 tok/s output (was 54.2)
Repo cleaned: INSTALL.bat + CHAT.bat + VERIFY.bat + MANIFEST.json + docs, archives on GitHub Releases

Known Issues — Please note this may have bugs

This is an experimental reference implementation. It works on my RX 6750 XT but could present bugs on other hardware.

Other RDNA2 cards (6600/6600 XT/6700 XT) not yet tested — may need 10.3.0 vs 10.3.1 or rocBLAS rebuild for gfx1030
enforce_eager=True required — torch.compile disabled on RDNA2 Windows
FP8 / AWQ not tested yet, multi-GPU not tested
If you test it on your RDNA2 card, please open an Issue with GPU model and logs — contributions welcome

GitHub Repo:

https://github.com/sebastianmechno-sys/vllm-rocm-windows-rdna2

Includes INSTALL.bat, CHAT.bat, VERIFY.bat, scripts/serve.py (OpenAI-compatible API), chat.html, assets with screenshots, docs/BUILD_ROCBLAS.md, MANIFEST.json with release checksums

Let me know if you test it on other RDNA2 cards.

Built on ROCm/TheRock, PyTorch ROCm, vLLM. Not affiliated with AMD. License Apache 2.0

I Got vLLM Running Natively on Windows on RX 6750 XT (RDNA2) – ROCm 7.15 PoC at 25.9 TFLOPS, No WSL2

Esseba-dev — Thu, 13 Aug 2026 21:13:10 +0000

I'm 17, mechanic from Torino. I built vLLM to run NATIVELY on Windows on my RX 6750 XT 12GB (gfx1031 / RDNA2).

AMD lists RDNA2 as "runtime-only" on ROCm 7.15 Windows. I got around it.

Repo PoC: https://github.com/sebastianmechno-sys/vllm-rocm-windows-rdna2
Upstream issue: https://github.com/ROCm/TheRock/issues/7341

The Problem

RX 6750 XT = gfx1031, not officially supported for full ROCm 7.15 Windows
vLLM depends on Triton attention kernels tuned for CDNA, they crash on RDNA2
FP8 not supported on RDNA2 (only FP16/INT8/INT4)

How I Fixed It

Built rocBLAS for gfx1031 via TheRock with HSA_OVERRIDE_GFX_VERSION=10.3.1
Custom ROCM_PATH + TheRock artifacts
vLLM 0.19 with custom plugin:


bash
VLLM_USE_TRITON_ATTN=0
VLLM_ROCM_USE_AITER=0
enforce_eager=True
torch.compile = off

Benchmarks (native Windows, no WSL2)
GEMM 4096x4096 FP16: 25.97 TFLOPS
opt-125m prompt: 107.9 tok/s, output: 54.2 tok/s (batch 8-32)
12GB VRAM: 10.3GB alloc, 5.47GB KV cache
FP16 stable, FP8 fails as expected on RDNA2
![ ](https://dev-to-uploads.s3.us-east-2.amazonaws.com/uploads/articles/hd9fxjwyruywcos112xw.png)
![ ](https://dev-to-uploads.s3.us-east-2.amazonaws.com/uploads/articles/a4p1357aiqgf9cm0w61w.png)

DEV Community: Esseba-dev