[WideEP] Remove pplx all2all backend (#33724)

Signed-off-by: Tyler Michael Smith <tlrmchlsmth@gmail.com>
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>

tests/kernels/moe/modular_kernel_tools/common.py

@@ -37,7 +37,6 @@ from vllm.utils.import_utils import (
     has_deep_ep,
     has_deep_gemm,
     has_mori,
-    has_pplx,
 )
 
 from .mk_objects import (
@@ -206,10 +205,6 @@ class Config:
         info = expert_info(self.fused_experts_type)
         return info.needs_deep_gemm
 
-    def needs_pplx(self):
-        info = prepare_finalize_info(self.prepare_finalize_type)
-        return info.backend == "pplx"
-
     def needs_deep_ep(self):
         info = prepare_finalize_info(self.prepare_finalize_type)
         return (
@@ -290,8 +285,6 @@ class Config:
             return False, "Needs DeepEP, but DeepEP not available."
         if self.needs_deep_gemm() and not has_deep_gemm():
             return False, "Needs DeepGEMM, but DeepGEMM not available."
-        if self.needs_pplx() and not has_pplx():  # noqa: SIM103
-            return False, "Needs PPLX, but PPLX not available."
         if self.needs_aiter() and not has_aiter():  # noqa: SIM103
             return False, "Needs Aiter, but Aiter not available."
         if self.needs_mori() and not has_mori():  # noqa: SIM103

tests/kernels/moe/modular_kernel_tools/mk_objects.py

@@ -39,7 +39,6 @@ from vllm.utils.import_utils import (
     has_deep_ep,
     has_deep_gemm,
     has_mori,
-    has_pplx,
 )
 
 
@@ -238,19 +237,6 @@ if has_mori():
         supports_apply_weight_on_input=False,
     )
 
-if has_pplx():
-    from vllm.model_executor.layers.fused_moe.pplx_prepare_finalize import (
-        PplxPrepareAndFinalize,
-    )
-
-    register_prepare_and_finalize(
-        PplxPrepareAndFinalize,
-        batched_format,
-        common_float_and_int_types,
-        blocked_quantization_support=True,
-        backend="pplx",
-    )
-
 if has_flashinfer_cutlass_fused_moe() and current_platform.has_device_capability(100):
     from vllm.model_executor.layers.fused_moe.flashinfer_a2a_prepare_finalize import (  # noqa: E501
         FlashInferCutlassMoEPrepareAndFinalize,

tests/kernels/moe/modular_kernel_tools/profile_modular_kernel.py

@@ -125,7 +125,7 @@ if __name__ == "__main__":
         description=(
             "Run single prepare-finalize & fused-experts combination test"
             "Example : python3 -m tests.kernels.moe.modular_kernel_tools.profile_modular_kernel "  # noqa: E501
-            "--pf-type PplxPrepareAndFinalize --experts-type BatchedTritonExperts"
+            "--pf-type DeepEPLLPrepareAndFinalize --experts-type BatchedTritonExperts"
         )
     )
     args = parser.parse_args()