HIP/ROCm fork optimized for AMD RDNA2 (gfx1030) with PrismML Q1_0_G128 1-bit quant support, RotorQuant, TurboQuant, EAGLE3 and P-EAGLE speculative decoding, and full Wave32 kernel optimizations.
-
Updated
Jun 23, 2026 - C++
#
gfx1030
Here are 9 public repositories matching this topic...
AMD ROCm (gfx1030) inference fork with RotorQuant/TurboQuant KV compression, PHANTOM-X zero-copy draft speculation, EAGLE3 speculative decoding, 12 RDNA2 crash fixes, and PrismML Bonsai Q1_0_G128 1-bit GGUF support.
triton hip bonsai rocm amd-gpu gguf speculative-decoding sglang rdna2 eagle3 turboquant prismml gfx1030 p-eagle radix-cache
-
Updated
Aug 6, 2026 - Python
Experimental open-source compatibility research for running DLSS Neural Rendering workloads on AMD RDNA2/gfx1030. Includes ISA translation, host emulation, proof-oriented validation, and bounded hardware testing. Early development: no end-to-end game frame yet.
open-source research compiler amd gpu reverse-engineering shader isa hip amdgpu rocm directx12 dlss neural-rendering gpu-compute navi21 rdna2 gfx1030 rx-6000
-
Updated
Sep 22, 2026 - Python
Exact FlashAttention for AMD RDNA2 (gfx1030 / RX 6800 XT), written in HIP. Drop-in replacement for flash-attn: 19.8 TFLOP/s forward, decode at 99% of HBM bandwidth.
-
Updated
Jul 20, 2026 - HIP
Custom SPIR-V kernel factory for PHANTOM speculative decoding — LLVM IR to GPU (SPIR-V/HIP) and CPU (native x86) cross-target compilation, RDNA2/gfx1030 optimized pre-compiled kernels, dynamic kernel swapping, zero-JIT inference pipeline
-
Updated
Apr 13, 2026 - LLVM
ComfyUI + ZLUDA on Windows for AMD RDNA2 desktop GPUs (RX 6950/6900/6800 XT, RX 6700/6750 XT) — fixes HIP/ZLUDA version mismatches, torch DLL loading, the mem-efficient attention backend that resets the display driver, and installs rocBLAS kernels on gfx1031.
windows rocm amd-gpu rocblas stable-diffusion comfyui rx6750xt rx6900xt rx6800xt zluda navi22 navi21 rdna2 gfx1031 rx6950xt gfx1030 rx6700xt hip-sdk
-
Updated
Sep 22, 2026 - PowerShell
Add this topic to your repo
To associate your repository with the gfx1030 topic, visit your repo's landing page and select "manage topics."