diff --git a/paddle/scripts/paddle_build.sh b/paddle/scripts/paddle_build.sh
index 3efca04a5c60acdd2267f867a57b90d35cefb002..fbe219a1c9cf85f19ae2ab991ae7e4207858f204 100755
--- a/paddle/scripts/paddle_build.sh
+++ b/paddle/scripts/paddle_build.sh
@@ -405,17 +405,19 @@ EOF
 
 function gen_dockerfile() {
     # Set BASE_IMAGE according to env variables
+    CUDA_MAJOR="$(echo $CUDA_VERSION | cut -d '.' -f 1).$(echo $CUDA_VERSION | cut -d '.' -f 2)"
+    CUDNN_MAJOR=$(echo $CUDNN_VERSION | cut -d '.' -f 1)
     if [[ ${WITH_GPU} == "ON" ]]; then
-    BASE_IMAGE="nvidia/cuda:8.0-cudnn5-runtime-ubuntu16.04"
+        BASE_IMAGE="nvidia/cuda:${CUDA_MAJOR}-cudnn${CUDNN_MAJOR}-runtime-ubuntu16.04"
     else
-    BASE_IMAGE="ubuntu:16.04"
+        BASE_IMAGE="ubuntu:16.04"
     fi
 
     DOCKERFILE_GPU_ENV=""
     DOCKERFILE_CUDNN_DSO=""
     if [[ ${WITH_GPU:-OFF} == 'ON' ]]; then
         DOCKERFILE_GPU_ENV="ENV LD_LIBRARY_PATH /usr/lib/x86_64-linux-gnu:\${LD_LIBRARY_PATH}"
-        DOCKERFILE_CUDNN_DSO="RUN ln -s /usr/lib/x86_64-linux-gnu/libcudnn.so.5 /usr/lib/x86_64-linux-gnu/libcudnn.so"
+        DOCKERFILE_CUDNN_DSO="RUN ln -s /usr/lib/x86_64-linux-gnu/libcudnn.so.${CUDNN_MAJOR} /usr/lib/x86_64-linux-gnu/libcudnn.so"
     fi
 
     cat <<EOF
@@ -490,7 +492,7 @@ function gen_fluid_inference_lib() {
     Deploying fluid inference library ...
     ========================================
 EOF
-        make inference_lib_dist
+        make -j `nproc` inference_lib_dist
     fi
 }