diff --git a/.github/scripts/fbgemm_gpu_build.bash b/.github/scripts/fbgemm_gpu_build.bash
index 1a7ca8826..254cd60ff 100644
--- a/.github/scripts/fbgemm_gpu_build.bash
+++ b/.github/scripts/fbgemm_gpu_build.bash
@@ -199,9 +199,9 @@ __configure_fbgemm_gpu_build_cuda () {
       #   https://github.com/NVIDIA/nvbench/discussions/129
       #   https://github.com/vllm-project/vllm/blob/main/CMakeLists.txt#L187
       #   https://github.com/NVIDIA/cutlass/blob/main/include/cutlass/gemm/kernel/sm90_gemm_tma_warpspecialized.hpp#L224
-      local arch_list="7.0;8.0;9.0;9.0a"
+      local arch_list="8.0;9.0;9.0a"
     else
-      local arch_list="7.0;8.0"
+      local arch_list="8.0"
     fi
   fi
   echo "[BUILD] Setting the following CUDA targets: ${arch_list}"
diff --git a/fbgemm_gpu/setup.py b/fbgemm_gpu/setup.py
index 53bd0377f..1ace81b39 100644
--- a/fbgemm_gpu/setup.py
+++ b/fbgemm_gpu/setup.py
@@ -247,7 +247,7 @@ def _get_cxx11_abi():
             _get_cxx11_abi(),
         ]
 
-        cxx_flags = ["-mcmodel=large", "-fuse-linker-plugin", "-flto"]
+        cxx_flags = ["-mcmodel=large", "-flto"]
 
         if self.args.verbose:
             print("[SETUP.PY] Building in VERBOSE mode ...")