From fdcc5176beb70e801d9baae14f992819b1ffcffb Mon Sep 17 00:00:00 2001
From: "Roberto L. Castro"
 <38211239+LopezCastroRoberto@users.noreply.github.com>
Date: Mon, 5 Jan 2026 21:11:35 +0100
Subject: [PATCH] [BugFix] Fix architecture flags to prevent issues on SM103
 (#31150)

Signed-off-by: LopezCastroRoberto <robertol.c510@gmail.com>
---
 cmake/external_projects/qutlass.cmake | 11 ++++++++---
 tools/flashinfer-build.sh             |  7 +++++--
 2 files changed, 13 insertions(+), 5 deletions(-)

diff --git a/cmake/external_projects/qutlass.cmake b/cmake/external_projects/qutlass.cmake
index 5a59a40999..84bb1b00c1 100644
--- a/cmake/external_projects/qutlass.cmake
+++ b/cmake/external_projects/qutlass.cmake
@@ -31,10 +31,15 @@ if(NOT qutlass_SOURCE_DIR)
 endif()
 message(STATUS "[QUTLASS] QuTLASS is available at ${qutlass_SOURCE_DIR}")
 
-cuda_archs_loose_intersection(QUTLASS_ARCHS "12.0a;10.0a" "${CUDA_ARCHS}")
-if(${CMAKE_CUDA_COMPILER_VERSION} VERSION_GREATER 12.8 AND QUTLASS_ARCHS)
+if(${CMAKE_CUDA_COMPILER_VERSION} VERSION_GREATER_EQUAL 13.0)
+  cuda_archs_loose_intersection(QUTLASS_ARCHS "12.0a;10.0f" "${CUDA_ARCHS}")
+else()
+  cuda_archs_loose_intersection(QUTLASS_ARCHS "12.0a;10.0a;10.3a" "${CUDA_ARCHS}")
+endif()
 
-  if(QUTLASS_ARCHS MATCHES "10\\.0a")
+if(${CMAKE_CUDA_COMPILER_VERSION} VERSION_GREATER_EQUAL 12.8 AND QUTLASS_ARCHS)
+
+  if(QUTLASS_ARCHS MATCHES "10\\.(0a|3a|0f)")
     set(QUTLASS_TARGET_CC 100)
   elseif(QUTLASS_ARCHS MATCHES "12\\.0a")
     set(QUTLASS_TARGET_CC 120)
diff --git a/tools/flashinfer-build.sh b/tools/flashinfer-build.sh
index 6c14d87348..b3cc6c3087 100755
--- a/tools/flashinfer-build.sh
+++ b/tools/flashinfer-build.sh
@@ -32,9 +32,12 @@ if [[ "${CUDA_VERSION}" == 11.* ]]; then
     FI_TORCH_CUDA_ARCH_LIST="7.5 8.0 8.9"
 elif [[ "${CUDA_VERSION}" == 12.[0-7]* ]]; then
     FI_TORCH_CUDA_ARCH_LIST="7.5 8.0 8.9 9.0a"
+elif [[ "${CUDA_VERSION}" == 12.[8-9]* ]]; then
+    # CUDA 12.8–12.9
+    FI_TORCH_CUDA_ARCH_LIST="7.5 8.0 8.9 9.0a 10.0a 10.3a 12.0"
 else
-    # CUDA 12.8+ supports 10.0a and 12.0
-    FI_TORCH_CUDA_ARCH_LIST="7.5 8.0 8.9 9.0a 10.0a 12.0"
+    # CUDA 13.0+
+    FI_TORCH_CUDA_ARCH_LIST="7.5 8.0 8.9 9.0a 10.0f 12.0"
 fi
 
 echo "🏗️ Building FlashInfer AOT for arches: ${FI_TORCH_CUDA_ARCH_LIST}"