[FlashInver v0.6.7] Integrate flashinfer_trtllm mxfp8 gemm (#21576)

python/sglang/srt/layers/quantization/fp8.py

@@ -495,20 +495,44 @@ class Fp8LinearMethod(LinearMethodBase):
             return
 
         if get_fp8_gemm_runner_backend().is_flashinfer_trtllm():
+            from flashinfer import shuffle_matrix_a, shuffle_matrix_sf_a
+
+            weight = layer.weight.data
+            scale_u8 = layer.weight_scale_inv.data
+            n, k = weight.shape
+            epilogue_tile_m = 128
+
+            copy_or_rebind_param(
+                layer,
+                "weight",
+                shuffle_matrix_a(
+                    weight.contiguous().view(torch.uint8), epilogue_tile_m
+                ).view(torch.float8_e4m3fn),
+            )
+            copy_or_rebind_param(
+                layer,
+                "weight_scale_inv",
+                shuffle_matrix_sf_a(
+                    scale_u8.contiguous().view(torch.uint8).reshape(n, k // 32),
+                    epilogue_tile_m,
+                    num_elts_per_sf=32,
+                )
+                .reshape_as(scale_u8)
+                .contiguous(),
+            )
+        elif get_fp8_gemm_runner_backend().is_flashinfer_cutlass():
             from flashinfer import block_scale_interleave
 
             scale_u8 = layer.weight_scale_inv.data
-            new_swizzled = block_scale_interleave(scale_u8.contiguous()).contiguous()
+            copy_or_rebind_param(
+                layer,
+                "weight_scale_inv",
+                block_scale_interleave(scale_u8.contiguous()).contiguous(),
+            )
         else:
             # Triton path consumes canonical 2D UE8M0 scales directly.
             return
 
-        copy_or_rebind_param(layer, "weight_scale_inv_swizzled", new_swizzled)
-        layer._weight_scale_inv_swizzled_src_version = layer.weight_scale_inv._version
-        layer._weight_scale_inv_swizzled_src_data_ptr = (
-            layer.weight_scale_inv.data_ptr()
-        )
-
     def _quantize_mxfp8_weights(self, layer: Module) -> None:
         weight = layer.weight.data
         qweight, weight_scale = mxfp8_group_quantize(weight)
@@ -657,22 +681,18 @@ class Fp8LinearMethod(LinearMethodBase):
             )
 
         if self.use_mxfp8:
-            if get_fp8_gemm_runner_backend().is_flashinfer_trtllm():
-                weight_scale = layer.weight_scale_inv_swizzled
-            else:
-                weight_scale = layer.weight_scale_inv
             if isinstance(x, tuple):
                 return self.w8a8_mxfp8_linear(
                     input=x[0],
                     weight=layer.weight,
-                    weight_scale=weight_scale,
+                    weight_scale=layer.weight_scale_inv,
                     input_scale=x[1],
                     bias=bias,
                 )
             return self.w8a8_mxfp8_linear(
                 input=x,
                 weight=layer.weight,
-                weight_scale=weight_scale,
+                weight_scale=layer.weight_scale_inv,
                 input_scale=None,
                 bias=bias,
             )

python/sglang/srt/layers/quantization/fp8_utils.py

@@ -214,6 +214,7 @@ def _check_cutlass_block_fp8_hardware_support() -> bool:
 
 
 if is_blackwell_supported() and is_flashinfer_available():
+    from flashinfer import SfLayout
     from flashinfer import mm_mxfp8 as _raw_flashinfer_mm_mxfp8
     from flashinfer import mxfp8_quantize as _raw_flashinfer_mxfp8_quantize
     from flashinfer.gemm import gemm_fp8_nt_groupwise as _raw_gemm_fp8_nt_groupwise
@@ -303,12 +304,13 @@ if is_blackwell_supported() and is_flashinfer_available():
             input,
             is_sf_swizzled_layout=is_sf_swizzled_layout,
             alignment=alignment,
+            sf_swizzle_layout=SfLayout.layout_128x4,
         )
 
     @register_custom_op(
         op_name="flashinfer_mm_mxfp8",
         mutates_args=[],
-        fake_impl=lambda q_input, weight_t, x_scale_u8, weight_scale_t, out_dtype, backend="auto": (
+        fake_impl=lambda q_input, weight_t, x_scale_u8, weight_scale_t, out_dtype, use_8x4_sf_layout=False, backend="auto": (
             q_input.new_empty((q_input.shape[0], weight_t.shape[1]), dtype=out_dtype)
         ),
     )
@@ -318,6 +320,7 @@ if is_blackwell_supported() and is_flashinfer_available():
         x_scale_u8: torch.Tensor,
         weight_scale_t: torch.Tensor,
         out_dtype: torch.dtype,
+        use_8x4_sf_layout: bool = False,
         backend: str = "auto",
     ) -> torch.Tensor:
         return _raw_flashinfer_mm_mxfp8(
@@ -326,6 +329,7 @@ if is_blackwell_supported() and is_flashinfer_available():
             x_scale_u8,
             weight_scale_t,
             out_dtype=out_dtype,
+            use_8x4_sf_layout=use_8x4_sf_layout,
             backend=backend,
         )
 
@@ -357,11 +361,13 @@ def dispatch_w8a8_mxfp8_linear() -> Callable:
     """Dispatch MXFP8 linear kernel by --fp8-gemm-backend.
 
     For MXFP8, Triton remains the default path. We only route to FlashInfer
-    when backend is explicitly set to flashinfer_trtllm.
+    when backend is explicitly set to flashinfer_cutlass or flashinfer_trtllm.
     """
     backend = get_fp8_gemm_runner_backend()
     if backend.is_flashinfer_trtllm():
         return flashinfer_mxfp8_blockscaled_linear
+    elif backend.is_flashinfer_cutlass():
+        return flashinfer_mxfp8_blockscaled_linear
     return triton_mxfp8_blockscaled_linear
 
 
@@ -962,6 +968,7 @@ def flashinfer_mxfp8_blockscaled_linear(
         )
     else:
         q_input = input_2d
+        x_scale_u8 = input_scale.contiguous()
 
     if output_dtype is None:
         if input_2d.dtype in (torch.float16, torch.bfloat16, torch.float32):
@@ -971,6 +978,20 @@ def flashinfer_mxfp8_blockscaled_linear(
 
     # Ensure transposed tensors are contiguous for FlashInfer's internal runner.
     weight_t = weight.contiguous().t()
+
+    if get_fp8_gemm_runner_backend().is_flashinfer_trtllm():
+
+        weight_scale_t = weight_scale.contiguous().view(-1)
+        output = flashinfer_mm_mxfp8(
+            q_input,
+            weight_t,
+            x_scale_u8,
+            weight_scale_t,
+            out_dtype=output_dtype,
+            use_8x4_sf_layout=False,
+            backend="trtllm",
+        )
+    elif get_fp8_gemm_runner_backend().is_flashinfer_cutlass():
         weight_scale_t = (
             weight_scale.contiguous().t()
             if weight_scale.ndim == 2
@@ -982,7 +1003,8 @@ def flashinfer_mxfp8_blockscaled_linear(
             x_scale_u8,
             weight_scale_t,
             out_dtype=output_dtype,
-        backend="auto",
+            use_8x4_sf_layout=False,
+            backend="cutlass",
         )
 
     if bias is not None:

test/registered/quant/test_fp8_blockwise_gemm.py

@@ -131,5 +131,10 @@ class TestMXFP8GemmFlashinferTrtllm(MXFP8GemmBase, unittest.TestCase):
     backend = "flashinfer_trtllm"
 
 
+@unittest.skipIf(get_device_sm() < 100, "Test requires CUDA SM 100 or higher")
+class TestMXFP8GemmFlashinferCutlass(MXFP8GemmBase, unittest.TestCase):
+    backend = "flashinfer_cutlass"
+
+
 if __name__ == "__main__":
     unittest.main()