From fc5bbc2dabadb80ff30fa7e5bb3318ac1a017d2f Mon Sep 17 00:00:00 2001
From: Gabriel Wu <13583761+lucifer1004@users.noreply.github.com>
Date: Fri, 27 Feb 2026 16:22:55 +0800
Subject: [PATCH] Fix typo in cute.nvgpu.warpgroup.mma doc (#2548)

---
 python/CuTeDSL/cutlass/cute/nvgpu/warpgroup/mma.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/python/CuTeDSL/cutlass/cute/nvgpu/warpgroup/mma.py b/python/CuTeDSL/cutlass/cute/nvgpu/warpgroup/mma.py
index f3c0ecb6..d2f89da6 100644
--- a/python/CuTeDSL/cutlass/cute/nvgpu/warpgroup/mma.py
+++ b/python/CuTeDSL/cutlass/cute/nvgpu/warpgroup/mma.py
@@ -333,7 +333,7 @@ class MmaF16BF16Trait(MmaTraits):
 @dataclass(frozen=True)
 class MmaF8Op(MmaOp):
     """
-    F16/BF16 warpgroup MMA Operation.
+    FP8 warpgroup MMA Operation.
 
     See the `PTX documentation <https://docs.nvidia.com/cuda/parallel-thread-execution/#asynchronous-warpgroup-level-matrix-instructions-wgmma-mma>`__.
     This Operation covers the instructions using the ``.e4m3`` or ``.e5m2`` qualifiers for the input operands.