diff --git a/CMakeLists.txt b/CMakeLists.txt
index 3b8bbd2e0272..c65e91bbec31 100644
--- a/CMakeLists.txt
+++ b/CMakeLists.txt
@@ -187,6 +187,7 @@ if(USE_TENSORRT)
   include_directories(${ONNX_PATH})
   include_directories(3rdparty/onnx-tensorrt/)
   include_directories(3rdparty/)
+  include_directories(3rdparty/onnx-tensorrt/third_party/onnx/)
   add_definitions(-DMXNET_USE_TENSORRT=1)
   add_definitions(-DONNX_NAMESPACE=onnx)
 
diff --git a/Jenkinsfile b/Jenkinsfile
index a7e3dbd9bcc0..89a2c34e3be5 100644
--- a/Jenkinsfile
+++ b/Jenkinsfile
@@ -60,7 +60,7 @@ mx_cmake_lib = 'build/libmxnet.so, build/libmxnet.a, build/3rdparty/dmlc-core/li
 mx_cmake_lib_debug = 'build/libmxnet.so, build/libmxnet.a, build/3rdparty/dmlc-core/libdmlc.a, build/tests/mxnet_unit_tests'
 mx_cmake_mkldnn_lib = 'build/libmxnet.so, build/libmxnet.a, build/3rdparty/dmlc-core/libdmlc.a, build/tests/mxnet_unit_tests, build/3rdparty/openmp/runtime/src/libomp.so, build/3rdparty/mkldnn/src/libmkldnn.so.0'
 mx_mkldnn_lib = 'lib/libmxnet.so, lib/libmxnet.a, lib/libiomp5.so, lib/libmkldnn.so.0, lib/libmklml_intel.so, 3rdparty/dmlc-core/libdmlc.a, 3rdparty/tvm/nnvm/lib/libnnvm.a'
-mx_tensorrt_lib = 'lib/libmxnet.so, lib/libnvonnxparser_runtime.so.0, lib/libnvonnxparser.so.0, lib/libonnx_proto.so, lib/libonnx.so'
+mx_tensorrt_lib = 'build/libmxnet.so, lib/libnvonnxparser_runtime.so.0, lib/libnvonnxparser.so.0, lib/libonnx_proto.so, lib/libonnx.so'
 mx_lib_cpp_examples = 'lib/libmxnet.so, lib/libmxnet.a, 3rdparty/dmlc-core/libdmlc.a, 3rdparty/tvm/nnvm/lib/libnnvm.a, 3rdparty/ps-lite/build/libps.a, deps/lib/libprotobuf-lite.a, deps/lib/libzmq.a, build/cpp-package/example/lenet, build/cpp-package/example/alexnet, build/cpp-package/example/googlenet, build/cpp-package/example/lenet_with_mxdataiter, build/cpp-package/example/resnet, build/cpp-package/example/mlp, build/cpp-package/example/mlp_cpu, build/cpp-package/example/mlp_gpu, build/cpp-package/example/test_score, build/cpp-package/example/test_optimizer'
 mx_lib_cpp_examples_cpu = 'build/libmxnet.so, build/cpp-package/example/mlp_cpu'
 
diff --git a/ci/docker/install/ubuntu_core.sh b/ci/docker/install/ubuntu_core.sh
index 64f8af3e0444..35430f2d3aa5 100755
--- a/ci/docker/install/ubuntu_core.sh
+++ b/ci/docker/install/ubuntu_core.sh
@@ -26,7 +26,6 @@ apt-get install -y \
     apt-transport-https \
     build-essential \
     ca-certificates \
-    cmake \
     curl \
     git \
     libatlas-base-dev \
@@ -41,3 +40,12 @@ apt-get install -y \
     sudo \
     unzip \
     wget
+
+# Note: we specify an exact cmake version to work around a cmake 3.10 CUDA 10 issue.
+# Reference: /~https://github.com/clab/dynet/issues/1457
+mkdir /opt/cmake && cd /opt/cmake
+wget -nv https://cmake.org/files/v3.12/cmake-3.12.4-Linux-x86_64.sh
+sh cmake-3.12.4-Linux-x86_64.sh --prefix=/opt/cmake --skip-license
+ln -s /opt/cmake/bin/cmake /usr/local/bin/cmake
+rm cmake-3.12.4-Linux-x86_64.sh
+cmake --version
diff --git a/ci/docker/runtime_functions.sh b/ci/docker/runtime_functions.sh
index 39631f9dc7e6..f2761331b246 100755
--- a/ci/docker/runtime_functions.sh
+++ b/ci/docker/runtime_functions.sh
@@ -521,22 +521,23 @@ build_ubuntu_gpu_tensorrt() {
     cp -L 3rdparty/onnx-tensorrt/build/libnvonnxparser_runtime.so.0 /work/mxnet/lib/
     cp -L 3rdparty/onnx-tensorrt/build/libnvonnxparser.so.0 /work/mxnet/lib/
 
-    rm -rf build
-    make \
-        DEV=1                                                \
-        ENABLE_TESTCOVERAGE=1                                \
-        USE_BLAS=openblas                                    \
-        USE_CUDA=1                                           \
-        USE_CUDA_PATH=/usr/local/cuda                        \
-        USE_CUDNN=1                                          \
-        USE_OPENCV=0                                         \
-        USE_DIST_KVSTORE=0                                   \
-        USE_TENSORRT=1                                       \
-        USE_JEMALLOC=0                                       \
-        USE_GPERFTOOLS=0                                     \
-        ONNX_NAMESPACE=onnx                                  \
-        CUDA_ARCH="-gencode arch=compute_70,code=compute_70" \
-        -j$(nproc)
+    cd /work/build
+    cmake -DUSE_CUDA=1                            \
+          -DCMAKE_CXX_COMPILER_LAUNCHER=ccache    \
+          -DCMAKE_C_COMPILER_LAUNCHER=ccache      \
+          -DUSE_CUDNN=1                           \
+          -DUSE_OPENCV=1                          \
+          -DUSE_TENSORRT=1                        \
+          -DUSE_OPENMP=0                          \
+          -DUSE_MKLDNN=0                          \
+          -DUSE_MKL_IF_AVAILABLE=OFF              \
+          -DENABLE_TESTCOVERAGE=ON                \
+          -DCUDA_ARCH_NAME=Manual                 \
+          -DCUDA_ARCH_BIN=$CI_CMAKE_CUDA_ARCH_BIN \
+          -G Ninja                                \
+          /work/mxnet
+
+    ninja -v
 }
 
 build_ubuntu_gpu_mkldnn() {