[TEST][ATen][CUDA] Skip row-wise scaled matrix mmultiplication tests on sm_120+ (pytorch#152814)

The float8 row-wise scaled matmuls are not supported on Blackwell yet. This PR adds skips to those tests to decrease the noise on `sm_120+` machines.

Pull Request resolved: pytorch#152814
Approved by: https://github.com/eqy, https://github.com/Skylion007

Author: Aidyn-A

test/test_matmul_cuda.py

@@ -24,6 +24,7 @@
     SM89OrLater,
     SM90OrLater,
     xfailIfSM100OrLater,
+    xfailIfSM120OrLater,
     _get_torch_cuda_version,
     PLATFORM_SUPPORTS_FP8,
     PLATFORM_SUPPORTS_MX_GEMM,
@@ -1012,8 +1013,9 @@ def test_float8_scale_fast_accum(self, device) -> None:
         out_fp8_s = torch._scaled_mm(x, y, scale_a=scale_a, scale_b=scale_b, use_fast_accum=True)
         self.assertEqual(out_fp8, out_fp8_s)
 
+    @xfailIfSM120OrLater
     @unittest.skipIf(not PLATFORM_SUPPORTS_FP8 or IS_WINDOWS, f8_msg)
-    @unittest.skipIf(not SM89OrLater, "rowwise implementation is currently sm89+ specific")
+    @unittest.skipIf(not SM89OrLater, "rowwise implementation is currently sm89-sm100 specific")
     @parametrize("use_fast_accum", [True, False])
     def test_float8_rowwise_scaling_sanity(self, device, use_fast_accum: bool) -> None:
         M, K, N = (1024, 512, 2048)
@@ -1117,8 +1119,9 @@ def test_float8_error_messages(self, device) -> None:
                 out_dtype=torch.bfloat16,
             )
 
+    @xfailIfSM120OrLater
     @unittest.skipIf(not PLATFORM_SUPPORTS_FP8 or IS_WINDOWS, f8_msg)
-    @unittest.skipIf(not SM89OrLater, "rowwise implementation is currently sm89+ specific")
+    @unittest.skipIf(not SM89OrLater, "rowwise implementation is currently sm89-sm100 specific")
     @parametrize("base_dtype", [torch.bfloat16])
     def test_scaled_mm_vs_emulated_row_wise(self, base_dtype):
         torch.manual_seed(42)

torch/testing/_internal/common_cuda.py

@@ -33,6 +33,7 @@
 SM89OrLater = LazyVal(lambda: torch.cuda.is_available() and torch.cuda.get_device_capability() >= (8, 9))
 SM90OrLater = LazyVal(lambda: torch.cuda.is_available() and torch.cuda.get_device_capability() >= (9, 0))
 SM100OrLater = LazyVal(lambda: torch.cuda.is_available() and torch.cuda.get_device_capability() >= (10, 0))
+SM120OrLater = LazyVal(lambda: torch.cuda.is_available() and torch.cuda.get_device_capability() >= (12, 0))
 
 IS_THOR = LazyVal(lambda: torch.cuda.is_available() and torch.cuda.get_device_capability()[0] == 10
                   and torch.cuda.get_device_capability()[1] > 0)
@@ -335,6 +336,9 @@ def xfailIfSM89(func):
 def xfailIfSM100OrLater(func):
     return func if not SM100OrLater else unittest.expectedFailure(func)
 
+def xfailIfSM120OrLater(func):
+    return func if not SM120OrLater else unittest.expectedFailure(func)
+
 def xfailIfDistributedNotSupported(func):
     return func if not (IS_MACOS or IS_JETSON) else unittest.expectedFailure(func)