From 817294a93e9aa6e15b3b106568552bd7a5bfa02d Mon Sep 17 00:00:00 2001
From: uvos <carl@uvos.xyz>
Date: Thu, 1 Oct 2026 07:51:38 +0200
Subject: [PATCH] HIP: avoid treating CDNA as dgx spark for gqa_ratio 20 in
 fattn_mma dqk 576 (llama/29572)

---
 ggml/src/ggml-cuda/fattn.cu | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/ggml/src/ggml-cuda/fattn.cu b/ggml/src/ggml-cuda/fattn.cu
index ad4d1df2c..156b93c00 100644
--- a/ggml/src/ggml-cuda/fattn.cu
+++ b/ggml/src/ggml-cuda/fattn.cu
@@ -365,7 +365,7 @@ static void ggml_cuda_flash_attn_ext_mma_f16(ggml_backend_cuda_context & ctx, gg
 
             GGML_ASSERT(Q->ne[2] % K->ne[2] == 0);
             const int gqa_ratio = Q->ne[2] / K->ne[2];
-            if (gqa_ratio == 20) { // GLM 4.7 Flash
+            if (gqa_ratio == 20 && GGML_CUDA_CC_IS_NVIDIA(cc)) { // GLM 4.7 Flash
                 if (cc >= GGML_CUDA_CC_DGX_SPARK) {
                     if (Q->ne[1] <= 8) {
                         ggml_cuda_flash_attn_ext_mma_f16_switch_ncols1<576, 512, 16>(ctx, dst);