diff --git a/paddle/memory/memory.cc b/paddle/memory/memory.cc
index 4056a54b4a3a35fdd32864f5d64d1bee7fda2160..78443cc35a400bceac77b99c3468daf16d8a4690 100644
--- a/paddle/memory/memory.cc
+++ b/paddle/memory/memory.cc
@@ -86,18 +86,22 @@ size_t Used<platform::GPUPlace>(platform::GPUPlace place) {
 }
 
 template <>
-void Copy<platform::CPUPlace, platform::GPUPlace>(platform::CPUPlace, void* dst,
-                                                  platform::GPUPlace,
+void Copy<platform::CPUPlace, platform::GPUPlace>(platform::CPUPlace dst_place,
+                                                  void* dst,
+                                                  platform::GPUPlace src_place,
                                                   const void* src, size_t num,
                                                   cudaStream_t stream) {
+  platform::SetDeviceId(src_place.device);
   platform::GpuMemcpyAsync(dst, src, num, cudaMemcpyDeviceToHost, stream);
 }
 
 template <>
-void Copy<platform::GPUPlace, platform::CPUPlace>(platform::GPUPlace, void* dst,
-                                                  platform::CPUPlace,
+void Copy<platform::GPUPlace, platform::CPUPlace>(platform::GPUPlace dst_place,
+                                                  void* dst,
+                                                  platform::CPUPlace src_place,
                                                   const void* src, size_t num,
                                                   cudaStream_t stream) {
+  platform::SetDeviceId(dst_place.device);
   platform::GpuMemcpyAsync(dst, src, num, cudaMemcpyHostToDevice, stream);
 }
 
@@ -108,6 +112,7 @@ void Copy<platform::GPUPlace, platform::GPUPlace>(platform::GPUPlace dst_place,
                                                   const void* src, size_t num,
                                                   cudaStream_t stream) {
   if (dst_place == src_place) {
+    platform::SetDeviceId(src_place.device);
     platform::GpuMemcpyAsync(dst, src, num, cudaMemcpyDeviceToDevice, stream);
   } else {
     platform::GpuMemcpyPeer(dst, dst_place.device, src, src_place.device, num,