fix

Author: OutisLi

source/CMakeLists.txt

@@ -356,6 +356,16 @@ if(ENABLE_PYTORCH AND NOT DEEPMD_C_ROOT)
     endif()
   endif()
   find_package(Torch REQUIRED)
+  # TorchConfig populates TORCH_CUDA_LIBRARIES only when PyTorch itself was
+  # built with CUDA. The fused graph-lower operators in op/pt include ATen CUDA
+  # headers and link libtorch_cuda, so they can be compiled only against a
+  # CUDA-enabled PyTorch: a CPU-only wheel ships the ATen/c10 CUDA headers but
+  # neither the generated c10/cuda/impl/cuda_cmake_macros.h nor libtorch_cuda.
+  if(TORCH_CUDA_LIBRARIES)
+    set(DEEPMD_TORCH_HAS_CUDA TRUE)
+  else()
+    set(DEEPMD_TORCH_HAS_CUDA FALSE)
+  endif()
   if(NOT Torch_VERSION VERSION_LESS "2.1.0")
     set_if_higher(CMAKE_CXX_STANDARD 17)
   elseif(NOT Torch_VERSION VERSION_LESS "1.5.0")

source/op/pt/CMakeLists.txt

@@ -1,9 +1,12 @@
 file(GLOB OP_SRC print_summary.cc comm.cc tabulate_multi_device.cc)
-# Fused graph-lower inference operators (CUDA / cuBLAS); GPU builds only. The
-# CUDA language is scoped per directory, so it must be enabled here: the sibling
-# GPU library turns it on only within its own subtree, which does not cover this
-# target.
-if(USE_CUDA_TOOLKIT)
+# Fused graph-lower inference operators (CUDA / cuBLAS). They include ATen CUDA
+# headers and link libtorch_cuda, so they build only against a CUDA-enabled
+# PyTorch (DEEPMD_TORCH_HAS_CUDA); against a CPU-only torch they are omitted and
+# the Python dispatch falls back to the reference path (see
+# deepmd.kernels.cuda.*.op_available). The CUDA language is scoped per
+# directory, so it must be enabled here: the sibling GPU library turns it on
+# only within its own subtree, which does not cover this target.
+if(USE_CUDA_TOOLKIT AND DEEPMD_TORCH_HAS_CUDA)
   find_package(CUDAToolkit REQUIRED)
   if(NOT DEFINED CMAKE_CUDA_ARCHITECTURES)
     # CUDA 12.9 and 13.x CCCL fail to compile CUB/Thrust with -arch=all.
@@ -28,7 +31,7 @@ endif()
 add_library(deepmd_op_pt MODULE ${OP_SRC})
 # link: libdeepmd libtorch
 target_link_libraries(deepmd_op_pt PRIVATE ${TORCH_LIBRARIES})
-if(USE_CUDA_TOOLKIT)
+if(USE_CUDA_TOOLKIT AND DEEPMD_TORCH_HAS_CUDA)
   target_link_libraries(deepmd_op_pt PRIVATE CUDA::cublas)
   # libtorch headers require C++17; the CUDA sources must match.
   set_target_properties(deepmd_op_pt PROPERTIES CUDA_STANDARD 17