From 065e523d7baa1beb67c9c6119c55f0ef203890fa Mon Sep 17 00:00:00 2001
From: fzyzcjy <5236035+fzyzcjy@users.noreply.github.com>
Date: Mon, 1 Sep 2025 14:29:56 +0800
Subject: [PATCH] Tiny allow DeepGEMM on cu12.9 (#9858)

---
 .../srt/layers/quantization/deep_gemm_wrapper/configurer.py    | 3 ---
 1 file changed, 3 deletions(-)

diff --git a/python/sglang/srt/layers/quantization/deep_gemm_wrapper/configurer.py b/python/sglang/srt/layers/quantization/deep_gemm_wrapper/configurer.py
index 936ca75b8..ecf7d1647 100644
--- a/python/sglang/srt/layers/quantization/deep_gemm_wrapper/configurer.py
+++ b/python/sglang/srt/layers/quantization/deep_gemm_wrapper/configurer.py
@@ -11,9 +11,6 @@ def _compute_enable_deep_gemm():
     sm_version = get_device_sm()
     if sm_version < 90:
         return False
-    # TODO fix deepgemm cu129 fp8 issue
-    if torch.version.cuda == "12.9":
-        return False
 
     try:
         import deep_gemm