[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>

tools/ep_kernels/README.md

@@ -4,7 +4,7 @@ Large-scale cluster-level expert parallel, as described in the [DeepSeek-V3 Tech
 
 Here we break down the requirements in 2 steps:
 
-1. Build and install the Python libraries (both [pplx-kernels](https://github.com/ppl-ai/pplx-kernels) and [DeepEP](https://github.com/deepseek-ai/DeepEP)), including necessary dependencies like NVSHMEM. This step does not require any privileged access. Any user can do this.
+1. Build and install the Python libraries ([DeepEP](https://github.com/deepseek-ai/DeepEP)), including necessary dependencies like NVSHMEM. This step does not require any privileged access. Any user can do this.
 2. Configure NVIDIA driver to enable IBGDA. This step requires root access, and must be done on the host machine.
 
 Step 2 is necessary for multi-node deployment.

tools/ep_kernels/elastic_ep/install_eep_libraries.sh

@@ -76,11 +76,4 @@ popd
 
 export CMAKE_PREFIX_PATH=$WORKSPACE/nvshmem_install:$CMAKE_PREFIX_PATH
 
-# build and install pplx, require pytorch installed
-pushd "$WORKSPACE"
-git clone https://github.com/ppl-ai/pplx-kernels
-cd pplx-kernels
-# see https://github.com/pypa/pip/issues/9955#issuecomment-838065925
-# PIP_NO_BUILD_ISOLATION=0 disables build isolation
-PIP_NO_BUILD_ISOLATION=0 TORCH_CUDA_ARCH_LIST=9.0a+PTX pip install . --no-deps -v
 

tools/ep_kernels/install_python_libraries.sh

@@ -4,12 +4,10 @@ set -ex
 # usage: ./install_python_libraries.sh [options]
 #   --workspace <dir>    workspace directory (default: ./ep_kernels_workspace)
 #   --mode <mode>        "install" (default) or "wheel"
-#   --pplx-ref <commit>  pplx-kernels commit hash
 #   --deepep-ref <commit> DeepEP commit hash
 #   --nvshmem-ver <ver>  NVSHMEM version 
 
 CUDA_HOME=${CUDA_HOME:-/usr/local/cuda}
-PPLX_COMMIT_HASH=${PPLX_COMMIT_HASH:-"12cecfd"}
 DEEPEP_COMMIT_HASH=${DEEPEP_COMMIT_HASH:-"73b6ea4"}
 NVSHMEM_VER=${NVSHMEM_VER:-"3.3.24"}  # Default supports both CUDA 12 and 13
 WORKSPACE=${WORKSPACE:-$(pwd)/ep_kernels_workspace}
@@ -35,14 +33,6 @@ while [[ $# -gt 0 ]]; do
             MODE="$2"
             shift 2
             ;;
-        --pplx-ref)
-            if [[ -z "$2" || "$2" =~ ^- ]]; then
-                echo "Error: --pplx-ref requires an argument." >&2
-                exit 1
-            fi
-            PPLX_COMMIT_HASH="$2"
-            shift 2
-            ;;
         --deepep-ref)
             if [[ -z "$2" || "$2" =~ ^- ]]; then
                 echo "Error: --deepep-ref requires an argument." >&2
@@ -188,14 +178,6 @@ do_build() {
     popd
 }
 
-# build pplx-kernels
-do_build \
-    "https://github.com/ppl-ai/pplx-kernels" \
-    "pplx-kernels" \
-    "setup.py" \
-    "$PPLX_COMMIT_HASH" \
-    ""
-
 # build DeepEP
 do_build \
     "https://github.com/deepseek-ai/DeepEP" \