diff --git a/paddle/fluid/platform/gpu_info.cc b/paddle/fluid/platform/gpu_info.cc
index add5cabd444bfe1cff259c1626835dbad2c8b63c..c8d312c61bdd7e6e420b69de417b2b8d69ab2b8a 100644
--- a/paddle/fluid/platform/gpu_info.cc
+++ b/paddle/fluid/platform/gpu_info.cc
@@ -32,6 +32,11 @@ constexpr static float fraction_reserve_gpu_memory = 0.05f;
 namespace paddle {
 namespace platform {
 
+/* Here is a very simple CUDA “pro tip”: cudaDeviceGetAttribute() is a much
+faster way to query device properties. You can see details in
+https://devblogs.nvidia.com/cuda-pro-tip-the-fast-way-to-query-device-properties/
+*/
+
 inline std::string CudaErrorWebsite() {
   return "Please see detail in https://docs.nvidia.com/cuda/cuda-runtime-api"
          "/group__CUDART__TYPES.html#group__CUDART__TYPES_1g3f51e3575c217824"
@@ -75,14 +80,23 @@ int GetCUDADeviceCount() {
 
 int GetCUDAComputeCapability(int id) {
   PADDLE_ENFORCE_LT(id, GetCUDADeviceCount(), "id must less than GPU count");
-  cudaDeviceProp device_prop;
-  auto error_code = cudaGetDeviceProperties(&device_prop, id);
-  PADDLE_ENFORCE(
-      error_code,
-      "cudaGetDeviceProperties failed in "
+  int major, minor;
+
+  auto major_error_code =
+      cudaDeviceGetAttribute(&major, cudaDevAttrComputeCapabilityMajor, id);
+  auto minor_error_code =
+      cudaDeviceGetAttribute(&minor, cudaDevAttrComputeCapabilityMinor, id);
+  PADDLE_ENFORCE_EQ(
+      major_error_code, 0,
+      "cudaDevAttrComputeCapabilityMajor failed in "
       "paddle::platform::GetCUDAComputeCapability, error code : %d, %s",
-      error_code, CudaErrorWebsite());
-  return device_prop.major * 10 + device_prop.minor;
+      major_error_code, CudaErrorWebsite());
+  PADDLE_ENFORCE_EQ(
+      minor_error_code, 0,
+      "cudaDevAttrComputeCapabilityMinor failed in "
+      "paddle::platform::GetCUDAComputeCapability, error code : %d, %s",
+      minor_error_code, CudaErrorWebsite());
+  return major * 10 + minor;
 }
 
 int GetCUDARuntimeVersion(int id) {