[Refactor] Optimize FP8 MOE Backend Choice and Log (#26044)

Signed-off-by: yewentao256 <zhyanwentao@126.com>

vllm/model_executor/layers/quantization/fp8.py

@@ -1,6 +1,7 @@
 # SPDX-License-Identifier: Apache-2.0
 # SPDX-FileCopyrightText: Copyright contributors to the vLLM project
 
+from enum import Enum
 from typing import TYPE_CHECKING, Any, Callable, Optional, Union
 
 import torch
@@ -68,6 +69,65 @@ ACTIVATION_SCHEMES = ["static", "dynamic"]
 logger = init_logger(__name__)
 
 
+class Fp8MoeBackend(Enum):
+    NONE = 0
+    FLASHINFER_TRTLLM = 1
+    FLASHINFER_CUTLASS = 2
+    DEEPGEMM = 3
+    CUTLASS_BLOCK_SCALED_GROUPED_GEMM = 4
+    MARLIN = 5
+    TRITON = 6
+
+
+def get_fp8_moe_backend(block_quant: bool) -> Fp8MoeBackend:
+    """
+    Select the primary FP8 MoE backend
+    Note: Shape-specific fallbacks may still occur at runtime.
+    """
+    # prefer FlashInfer backends when available and enabled on supported GPUs
+    if (current_platform.is_cuda()
+            and current_platform.is_device_capability(100)
+            and envs.VLLM_USE_FLASHINFER_MOE_FP8 and has_flashinfer_moe()):
+        backend = get_flashinfer_moe_backend()
+        if backend == FlashinferMoeBackend.TENSORRT_LLM:
+            logger.info_once(
+                "Using FlashInfer FP8 MoE TRTLLM backend for SM100")
+            return Fp8MoeBackend.FLASHINFER_TRTLLM
+        else:
+            logger.info_once(
+                "Using FlashInfer FP8 MoE CUTLASS backend for SM100")
+            return Fp8MoeBackend.FLASHINFER_CUTLASS
+
+    # weight-only path for older GPUs without native FP8
+    use_marlin = (not current_platform.has_device_capability(89)
+                  or envs.VLLM_TEST_FORCE_FP8_MARLIN)
+    if current_platform.is_rocm():
+        use_marlin = False
+    if use_marlin:
+        logger.info_once("Using Marlin backend for FP8 MoE")
+        return Fp8MoeBackend.MARLIN
+
+    # deepGEMM on supported platforms with block-quantized weights
+    if envs.VLLM_USE_DEEP_GEMM and block_quant:
+        if not has_deep_gemm():
+            logger.warning_once(
+                "DeepGEMM backend requested but not available.")
+        elif is_deep_gemm_supported():
+            logger.info_once("Using DeepGEMM backend for FP8 MoE")
+            return Fp8MoeBackend.DEEPGEMM
+
+    # CUTLASS BlockScaled GroupedGemm on SM100 with block-quantized weights
+    if (current_platform.is_cuda()
+            and current_platform.is_device_capability(100) and block_quant):
+        logger.info_once(
+            "Using Cutlass BlockScaled GroupedGemm backend for FP8 MoE")
+        return Fp8MoeBackend.CUTLASS_BLOCK_SCALED_GROUPED_GEMM
+
+    # default to Triton
+    logger.info_once("Using Triton backend for FP8 MoE")
+    return Fp8MoeBackend.TRITON
+
+
 class Fp8Config(QuantizationConfig):
     """Config class for FP8."""
 
@@ -453,54 +513,19 @@ class Fp8MoEMethod(FusedMoEMethodBase):
         self.fused_experts: Optional[
             mk.FusedMoEModularKernel] = None  # type: ignore
 
-        # For GPUs that lack FP8 hardware support, we can leverage the Marlin
+        self.fp8_backend = get_fp8_moe_backend(self.block_quant)
-        # kernel for fast weight-only FP8 quantization
-        self.use_marlin = (not current_platform.has_device_capability(89)
-                           or envs.VLLM_TEST_FORCE_FP8_MARLIN)
-        # Disable marlin for rocm
-        if current_platform.is_rocm():
-            self.use_marlin = False
 
-        # First check for Flashinfer MOE on Blackwell GPUs
+        self.use_marlin = (self.fp8_backend == Fp8MoeBackend.MARLIN)
         self.flashinfer_moe_backend: Optional[FlashinferMoeBackend] = None
-        if (current_platform.is_cuda()
+        if self.fp8_backend == Fp8MoeBackend.FLASHINFER_TRTLLM:
-                and current_platform.is_device_capability(100)
+            self.flashinfer_moe_backend = FlashinferMoeBackend.TENSORRT_LLM
-                and envs.VLLM_USE_FLASHINFER_MOE_FP8 and has_flashinfer_moe()):
+        elif self.fp8_backend == Fp8MoeBackend.FLASHINFER_CUTLASS:
-            self.flashinfer_moe_backend = get_flashinfer_moe_backend()
+            self.flashinfer_moe_backend = FlashinferMoeBackend.CUTLASS
-            logger.info_once(
-                f"Detected Blackwell GPUs, using FlashInfer "
-                f"{self.flashinfer_moe_backend.value} kernels for FP8 MOE.")
 
-        # Check for DeepGemm support.
+        self.allow_deep_gemm = (self.fp8_backend == Fp8MoeBackend.DEEPGEMM)
-        self.allow_deep_gemm = False
+        self.allow_cutlass_block_scaled_grouped_gemm = (
-        if envs.VLLM_USE_DEEP_GEMM:
+            self.fp8_backend == Fp8MoeBackend.CUTLASS_BLOCK_SCALED_GROUPED_GEMM
-            if not has_deep_gemm():
+        )
-                logger.warning_once("Failed to import DeepGemm kernels.")
-            elif not self.block_quant:
-                logger.warning_once("Model is not block quantized. Not using"
-                                    " DeepGemm kernels")
-            elif self.flashinfer_moe_backend:
-                logger.info_once("DeepGemm disabled: FlashInfer MOE is"
-                                 " enabled.")
-            elif (is_deep_gemm_supported()):
-                logger.debug_once(
-                    "DeepGemm kernels available for Fp8MoEMethod.")
-                self.allow_deep_gemm = True
-            else:
-                logger.warning_once(
-                    "DeepGemm not supported on the current platform.")
-
-        # Check for CutlassBlockScaledGroupedGemm support.
-        self.allow_cutlass_block_scaled_grouped_gemm = False
-        if not self.block_quant:
-            logger.debug_once("Model is not block quantized. Not using "
-                              "CutlassBlockScaledGroupedGemm kernels")
-        elif (current_platform.is_cuda()
-              and current_platform.is_device_capability(100)
-              and not self.flashinfer_moe_backend):
-            logger.debug_once(
-                "CutlassBlockScaledGroupedGemm available for Fp8MoEMethod.")
-            self.allow_cutlass_block_scaled_grouped_gemm = True
 
     def create_weights(self, layer: Module, num_experts: int, hidden_size: int,
                        intermediate_size_per_partition: int,