diff --git a/tensorflow/tools/dockerfiles/assembler.py b/tensorflow/tools/dockerfiles/assembler.py
index 09537b73144..83b72cb5bb8 100644
--- a/tensorflow/tools/dockerfiles/assembler.py
+++ b/tensorflow/tools/dockerfiles/assembler.py
@@ -34,6 +34,7 @@ import errno
 import itertools
 import multiprocessing
 import os
+import platform
 import re
 import shutil
 import sys
@@ -552,6 +553,13 @@ def main(argv):
       if not FLAGS.build_images:
         continue
 
+      # Only build images for host architecture
+      proc_arch = platform.processor()
+      is_x86 = proc_arch.startswith('x86')
+      if (is_x86 and any([arch in tag for arch in ['ppc64le']]) or
+          not is_x86 and proc_arch not in tag):
+        continue
+
       # Generate a temporary Dockerfile to use to build, since docker-py
       # needs a filepath relative to the build context (i.e. the current
       # directory)
diff --git a/tensorflow/tools/dockerfiles/dockerfiles/devel-cpu-jupyter.Dockerfile b/tensorflow/tools/dockerfiles/dockerfiles/devel-cpu-jupyter.Dockerfile
index c1f6dafbe0b..4657f8b4c74 100644
--- a/tensorflow/tools/dockerfiles/dockerfiles/devel-cpu-jupyter.Dockerfile
+++ b/tensorflow/tools/dockerfiles/dockerfiles/devel-cpu-jupyter.Dockerfile
@@ -30,7 +30,6 @@ RUN apt-get update && apt-get install -y --no-install-recommends \
         libcurl3-dev \
         libfreetype6-dev \
         libhdf5-serial-dev \
-        libpng12-dev \
         libzmq3-dev \
         pkg-config \
         rsync \
@@ -43,12 +42,14 @@ RUN apt-get update && apt-get install -y --no-install-recommends \
         && \
     apt-get clean && \
     rm -rf /var/lib/apt/lists/*
- 
+
 ENV CI_BUILD_PYTHON python
 
-# Check out TensorFlow source code if --build_arg CHECKOUT_TENSORFLOW=1
+# CACHE_STOP is used to rerun future commands, otherwise cloning tensorflow will be cached and will not pull the most recent version
+ARG CACHE_STOP=1
+# Check out TensorFlow source code if --build_arg CHECKOUT_TF_SRC=1
 ARG CHECKOUT_TF_SRC=0
-RUN test "${CHECKOUT_TF_SRC}" -eq 1 && git clone https://github.com/tensorflow/tensorflow.git /tensorflow_src
+RUN test "${CHECKOUT_TF_SRC}" -eq 1 && git clone https://github.com/tensorflow/tensorflow.git /tensorflow_src || true
 
 ARG USE_PYTHON_3_NOT_2
 ARG _PY_SUFFIX=${USE_PYTHON_3_NOT_2:+3}
diff --git a/tensorflow/tools/dockerfiles/dockerfiles/devel-cpu.Dockerfile b/tensorflow/tools/dockerfiles/dockerfiles/devel-cpu.Dockerfile
index b4dfc8b0997..ce69ee0650c 100644
--- a/tensorflow/tools/dockerfiles/dockerfiles/devel-cpu.Dockerfile
+++ b/tensorflow/tools/dockerfiles/dockerfiles/devel-cpu.Dockerfile
@@ -30,7 +30,6 @@ RUN apt-get update && apt-get install -y --no-install-recommends \
         libcurl3-dev \
         libfreetype6-dev \
         libhdf5-serial-dev \
-        libpng12-dev \
         libzmq3-dev \
         pkg-config \
         rsync \
@@ -43,12 +42,14 @@ RUN apt-get update && apt-get install -y --no-install-recommends \
         && \
     apt-get clean && \
     rm -rf /var/lib/apt/lists/*
- 
+
 ENV CI_BUILD_PYTHON python
 
-# Check out TensorFlow source code if --build_arg CHECKOUT_TENSORFLOW=1
+# CACHE_STOP is used to rerun future commands, otherwise cloning tensorflow will be cached and will not pull the most recent version
+ARG CACHE_STOP=1
+# Check out TensorFlow source code if --build_arg CHECKOUT_TF_SRC=1
 ARG CHECKOUT_TF_SRC=0
-RUN test "${CHECKOUT_TF_SRC}" -eq 1 && git clone https://github.com/tensorflow/tensorflow.git /tensorflow_src
+RUN test "${CHECKOUT_TF_SRC}" -eq 1 && git clone https://github.com/tensorflow/tensorflow.git /tensorflow_src || true
 
 ARG USE_PYTHON_3_NOT_2
 ARG _PY_SUFFIX=${USE_PYTHON_3_NOT_2:+3}
diff --git a/tensorflow/tools/dockerfiles/dockerfiles/devel-gpu-jupyter.Dockerfile b/tensorflow/tools/dockerfiles/dockerfiles/devel-gpu-jupyter.Dockerfile
index 6d76c06332b..e41fc18e42b 100644
--- a/tensorflow/tools/dockerfiles/dockerfiles/devel-gpu-jupyter.Dockerfile
+++ b/tensorflow/tools/dockerfiles/dockerfiles/devel-gpu-jupyter.Dockerfile
@@ -21,23 +21,28 @@
 
 ARG UBUNTU_VERSION=16.04
 
-FROM nvidia/cuda:10.0-base-ubuntu${UBUNTU_VERSION} as base
+ARG ARCH=
+ARG CUDA=10.0
+FROM nvidia/cuda${ARCH:+-$ARCH}:${CUDA}-base-ubuntu${UBUNTU_VERSION} as base
+ARG CUDNN=7.4.1.5-1
+ARG LIB_DIR_PREFIX=x84_64
 
+# Needed for string substitution 
+SHELL ["/bin/bash", "-c"]
 RUN apt-get update && apt-get install -y --no-install-recommends \
         build-essential \
-        cuda-command-line-tools-10-0 \
-        cuda-cublas-dev-10-0 \
-        cuda-cudart-dev-10-0 \
-        cuda-cufft-dev-10-0 \
-        cuda-curand-dev-10-0 \
-        cuda-cusolver-dev-10-0 \
-        cuda-cusparse-dev-10-0 \
-        libcudnn7=7.4.1.5-1+cuda10.0 \
-        libcudnn7-dev=7.4.1.5-1+cuda10.0 \
+        cuda-command-line-tools-${CUDA/./-} \
+        cuda-cublas-dev-${CUDA/./-} \
+        cuda-cudart-dev-${CUDA/./-} \
+        cuda-cufft-dev-${CUDA/./-} \
+        cuda-curand-dev-${CUDA/./-} \
+        cuda-cusolver-dev-${CUDA/./-} \
+        cuda-cusparse-dev-${CUDA/./-} \
+        libcudnn7=${CUDNN}+cuda${CUDA} \
+        libcudnn7-dev=${CUDNN}+cuda${CUDA} \
         libcurl3-dev \
         libfreetype6-dev \
         libhdf5-serial-dev \
-        libpng12-dev \
         libzmq3-dev \
         pkg-config \
         rsync \
@@ -48,14 +53,15 @@ RUN apt-get update && apt-get install -y --no-install-recommends \
         wget \
         git \
         && \
-    find /usr/local/cuda-10.0/lib64/ -type f -name 'lib*_static.a' -not -name 'libcudart_static.a' -delete && \
-    rm /usr/lib/x86_64-linux-gnu/libcudnn_static_v7.a
+    find /usr/local/cuda-${CUDA}/lib64/ -type f -name 'lib*_static.a' -not -name 'libcudart_static.a' -delete && \
+    rm /usr/lib/${LIB_DIR_PREFIX}-linux-gnu/libcudnn_static_v7.a
 
-RUN apt-get update && \
-        apt-get install nvinfer-runtime-trt-repo-ubuntu1604-5.0.2-ga-cuda10.0 \
+RUN [ ${ARCH} = ppc64le ] || (apt-get update && \
+        apt-get install nvinfer-runtime-trt-repo-ubuntu1604-5.0.2-ga-cuda${CUDA} \
         && apt-get update \
-        && apt-get install -y --no-install-recommends libnvinfer-dev=5.0.2-1+cuda10.0 \
-        && rm -rf /var/lib/apt/lists/*
+        && apt-get install -y --no-install-recommends libnvinfer5=5.0.2-1+cuda${CUDA} \
+        && apt-get clean \
+        && rm -rf /var/lib/apt/lists/*)
 
 # Configure the build for our CUDA configuration.
 ENV CI_BUILD_PYTHON python
@@ -63,12 +69,13 @@ ENV LD_LIBRARY_PATH /usr/local/cuda/extras/CUPTI/lib64:$LD_LIBRARY_PATH
 ENV TF_NEED_CUDA 1
 ENV TF_NEED_TENSORRT 1
 ENV TF_CUDA_COMPUTE_CAPABILITIES=3.5,5.2,6.0,6.1,7.0
-ENV TF_CUDA_VERSION=10.0
-ENV TF_CUDNN_VERSION=7
-
-# Check out TensorFlow source code if --build_arg CHECKOUT_TENSORFLOW=1
+ENV TF_CUDA_VERSION=${CUDA}
+ENV TF_CUDNN_VERSION=${CUDNN%%.*}
+# CACHE_STOP is used to rerun future commands, otherwise cloning tensorflow will be cached and will not pull the most recent version
+ARG CACHE_STOP=1
+# Check out TensorFlow source code if --build_arg CHECKOUT_TF_SRC=1
 ARG CHECKOUT_TF_SRC=0
-RUN test "${CHECKOUT_TF_SRC}" -eq 1 && git clone https://github.com/tensorflow/tensorflow.git /tensorflow_src
+RUN test "${CHECKOUT_TF_SRC}" -eq 1 && git clone https://github.com/tensorflow/tensorflow.git /tensorflow_src || true
 
 ARG USE_PYTHON_3_NOT_2
 ARG _PY_SUFFIX=${USE_PYTHON_3_NOT_2:+3}
diff --git a/tensorflow/tools/dockerfiles/dockerfiles/devel-gpu.Dockerfile b/tensorflow/tools/dockerfiles/dockerfiles/devel-gpu.Dockerfile
index 160abc87639..7ae50100791 100644
--- a/tensorflow/tools/dockerfiles/dockerfiles/devel-gpu.Dockerfile
+++ b/tensorflow/tools/dockerfiles/dockerfiles/devel-gpu.Dockerfile
@@ -21,23 +21,28 @@
 
 ARG UBUNTU_VERSION=16.04
 
-FROM nvidia/cuda:10.0-base-ubuntu${UBUNTU_VERSION} as base
+ARG ARCH=
+ARG CUDA=10.0
+FROM nvidia/cuda${ARCH:+-$ARCH}:${CUDA}-base-ubuntu${UBUNTU_VERSION} as base
+ARG CUDNN=7.4.1.5-1
+ARG LIB_DIR_PREFIX=x84_64
 
+# Needed for string substitution 
+SHELL ["/bin/bash", "-c"]
 RUN apt-get update && apt-get install -y --no-install-recommends \
         build-essential \
-        cuda-command-line-tools-10-0 \
-        cuda-cublas-dev-10-0 \
-        cuda-cudart-dev-10-0 \
-        cuda-cufft-dev-10-0 \
-        cuda-curand-dev-10-0 \
-        cuda-cusolver-dev-10-0 \
-        cuda-cusparse-dev-10-0 \
-        libcudnn7=7.4.1.5-1+cuda10.0 \
-        libcudnn7-dev=7.4.1.5-1+cuda10.0 \
+        cuda-command-line-tools-${CUDA/./-} \
+        cuda-cublas-dev-${CUDA/./-} \
+        cuda-cudart-dev-${CUDA/./-} \
+        cuda-cufft-dev-${CUDA/./-} \
+        cuda-curand-dev-${CUDA/./-} \
+        cuda-cusolver-dev-${CUDA/./-} \
+        cuda-cusparse-dev-${CUDA/./-} \
+        libcudnn7=${CUDNN}+cuda${CUDA} \
+        libcudnn7-dev=${CUDNN}+cuda${CUDA} \
         libcurl3-dev \
         libfreetype6-dev \
         libhdf5-serial-dev \
-        libpng12-dev \
         libzmq3-dev \
         pkg-config \
         rsync \
@@ -48,14 +53,15 @@ RUN apt-get update && apt-get install -y --no-install-recommends \
         wget \
         git \
         && \
-    find /usr/local/cuda-10.0/lib64/ -type f -name 'lib*_static.a' -not -name 'libcudart_static.a' -delete && \
-    rm /usr/lib/x86_64-linux-gnu/libcudnn_static_v7.a
+    find /usr/local/cuda-${CUDA}/lib64/ -type f -name 'lib*_static.a' -not -name 'libcudart_static.a' -delete && \
+    rm /usr/lib/${LIB_DIR_PREFIX}-linux-gnu/libcudnn_static_v7.a
 
-RUN apt-get update && \
-        apt-get install nvinfer-runtime-trt-repo-ubuntu1604-5.0.2-ga-cuda10.0 \
+RUN [ ${ARCH} = ppc64le ] || (apt-get update && \
+        apt-get install nvinfer-runtime-trt-repo-ubuntu1604-5.0.2-ga-cuda${CUDA} \
         && apt-get update \
-        && apt-get install -y --no-install-recommends libnvinfer-dev=5.0.2-1+cuda10.0 \
-        && rm -rf /var/lib/apt/lists/*
+        && apt-get install -y --no-install-recommends libnvinfer5=5.0.2-1+cuda${CUDA} \
+        && apt-get clean \
+        && rm -rf /var/lib/apt/lists/*)
 
 # Configure the build for our CUDA configuration.
 ENV CI_BUILD_PYTHON python
@@ -63,12 +69,13 @@ ENV LD_LIBRARY_PATH /usr/local/cuda/extras/CUPTI/lib64:$LD_LIBRARY_PATH
 ENV TF_NEED_CUDA 1
 ENV TF_NEED_TENSORRT 1
 ENV TF_CUDA_COMPUTE_CAPABILITIES=3.5,5.2,6.0,6.1,7.0
-ENV TF_CUDA_VERSION=10.0
-ENV TF_CUDNN_VERSION=7
-
-# Check out TensorFlow source code if --build_arg CHECKOUT_TENSORFLOW=1
+ENV TF_CUDA_VERSION=${CUDA}
+ENV TF_CUDNN_VERSION=${CUDNN%%.*}
+# CACHE_STOP is used to rerun future commands, otherwise cloning tensorflow will be cached and will not pull the most recent version
+ARG CACHE_STOP=1
+# Check out TensorFlow source code if --build_arg CHECKOUT_TF_SRC=1
 ARG CHECKOUT_TF_SRC=0
-RUN test "${CHECKOUT_TF_SRC}" -eq 1 && git clone https://github.com/tensorflow/tensorflow.git /tensorflow_src
+RUN test "${CHECKOUT_TF_SRC}" -eq 1 && git clone https://github.com/tensorflow/tensorflow.git /tensorflow_src || true
 
 ARG USE_PYTHON_3_NOT_2
 ARG _PY_SUFFIX=${USE_PYTHON_3_NOT_2:+3}
diff --git a/tensorflow/tools/dockerfiles/dockerfiles/gpu-jupyter.Dockerfile b/tensorflow/tools/dockerfiles/dockerfiles/gpu-jupyter.Dockerfile
index 46252c54138..12eb5afa8af 100644
--- a/tensorflow/tools/dockerfiles/dockerfiles/gpu-jupyter.Dockerfile
+++ b/tensorflow/tools/dockerfiles/dockerfiles/gpu-jupyter.Dockerfile
@@ -21,32 +21,37 @@
 
 ARG UBUNTU_VERSION=16.04
 
-FROM nvidia/cuda:10.0-base-ubuntu${UBUNTU_VERSION} as base
+ARG ARCH=
+ARG CUDA=10.0
+FROM nvidia/cuda${ARCH:+-$ARCH}:${CUDA}-base-ubuntu${UBUNTU_VERSION} as base
+ARG CUDNN=7.4.1.5-1
 
+# Needed for string substitution 
+SHELL ["/bin/bash", "-c"]
 # Pick up some TF dependencies
 RUN apt-get update && apt-get install -y --no-install-recommends \
         build-essential \
-        cuda-command-line-tools-10-0 \
-        cuda-cublas-10-0 \
-        cuda-cufft-10-0 \
-        cuda-curand-10-0 \
-        cuda-cusolver-10-0 \
-        cuda-cusparse-10-0 \
-        libcudnn7=7.4.1.5-1+cuda10.0 \
+        cuda-command-line-tools-${CUDA/./-} \
+        cuda-cublas-${CUDA/./-} \
+        cuda-cufft-${CUDA/./-} \
+        cuda-curand-${CUDA/./-} \
+        cuda-cusolver-${CUDA/./-} \
+        cuda-cusparse-${CUDA/./-} \
+        curl \
+        libcudnn7=${CUDNN}+cuda${CUDA} \
         libfreetype6-dev \
         libhdf5-serial-dev \
-        libpng12-dev \
         libzmq3-dev \
         pkg-config \
         software-properties-common \
         unzip
 
-RUN apt-get update && \
-        apt-get install nvinfer-runtime-trt-repo-ubuntu1604-5.0.2-ga-cuda10.0 \
+RUN [ ${ARCH} = ppc64le ] || (apt-get update && \
+        apt-get install nvinfer-runtime-trt-repo-ubuntu1604-5.0.2-ga-cuda${CUDA} \
         && apt-get update \
-        && apt-get install -y --no-install-recommends libnvinfer5=5.0.2-1+cuda10.0 \
+        && apt-get install -y --no-install-recommends libnvinfer5=5.0.2-1+cuda${CUDA} \
         && apt-get clean \
-        && rm -rf /var/lib/apt/lists/*
+        && rm -rf /var/lib/apt/lists/*)
 
 # For CUDA profiling, TensorFlow requires CUPTI.
 ENV LD_LIBRARY_PATH /usr/local/cuda/extras/CUPTI/lib64:$LD_LIBRARY_PATH
diff --git a/tensorflow/tools/dockerfiles/dockerfiles/gpu.Dockerfile b/tensorflow/tools/dockerfiles/dockerfiles/gpu.Dockerfile
index 80e427f824a..00664b6b73f 100644
--- a/tensorflow/tools/dockerfiles/dockerfiles/gpu.Dockerfile
+++ b/tensorflow/tools/dockerfiles/dockerfiles/gpu.Dockerfile
@@ -21,32 +21,37 @@
 
 ARG UBUNTU_VERSION=16.04
 
-FROM nvidia/cuda:10.0-base-ubuntu${UBUNTU_VERSION} as base
+ARG ARCH=
+ARG CUDA=10.0
+FROM nvidia/cuda${ARCH:+-$ARCH}:${CUDA}-base-ubuntu${UBUNTU_VERSION} as base
+ARG CUDNN=7.4.1.5-1
 
+# Needed for string substitution 
+SHELL ["/bin/bash", "-c"]
 # Pick up some TF dependencies
 RUN apt-get update && apt-get install -y --no-install-recommends \
         build-essential \
-        cuda-command-line-tools-10-0 \
-        cuda-cublas-10-0 \
-        cuda-cufft-10-0 \
-        cuda-curand-10-0 \
-        cuda-cusolver-10-0 \
-        cuda-cusparse-10-0 \
-        libcudnn7=7.4.1.5-1+cuda10.0 \
+        cuda-command-line-tools-${CUDA/./-} \
+        cuda-cublas-${CUDA/./-} \
+        cuda-cufft-${CUDA/./-} \
+        cuda-curand-${CUDA/./-} \
+        cuda-cusolver-${CUDA/./-} \
+        cuda-cusparse-${CUDA/./-} \
+        curl \
+        libcudnn7=${CUDNN}+cuda${CUDA} \
         libfreetype6-dev \
         libhdf5-serial-dev \
-        libpng12-dev \
         libzmq3-dev \
         pkg-config \
         software-properties-common \
         unzip
 
-RUN apt-get update && \
-        apt-get install nvinfer-runtime-trt-repo-ubuntu1604-5.0.2-ga-cuda10.0 \
+RUN [ ${ARCH} = ppc64le ] || (apt-get update && \
+        apt-get install nvinfer-runtime-trt-repo-ubuntu1604-5.0.2-ga-cuda${CUDA} \
         && apt-get update \
-        && apt-get install -y --no-install-recommends libnvinfer5=5.0.2-1+cuda10.0 \
+        && apt-get install -y --no-install-recommends libnvinfer5=5.0.2-1+cuda${CUDA} \
         && apt-get clean \
-        && rm -rf /var/lib/apt/lists/*
+        && rm -rf /var/lib/apt/lists/*)
 
 # For CUDA profiling, TensorFlow requires CUPTI.
 ENV LD_LIBRARY_PATH /usr/local/cuda/extras/CUPTI/lib64:$LD_LIBRARY_PATH
diff --git a/tensorflow/tools/dockerfiles/dockerfiles/ppc64le/cpu-ppc64le-jupyter.Dockerfile b/tensorflow/tools/dockerfiles/dockerfiles/ppc64le/cpu-ppc64le-jupyter.Dockerfile
new file mode 100644
index 00000000000..beb3292a9db
--- /dev/null
+++ b/tensorflow/tools/dockerfiles/dockerfiles/ppc64le/cpu-ppc64le-jupyter.Dockerfile
@@ -0,0 +1,92 @@
+# Copyright 2018 The TensorFlow Authors. All Rights Reserved.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+# ============================================================================
+#
+# THIS IS A GENERATED DOCKERFILE.
+#
+# This file was assembled from multiple pieces, whose use is documented
+# throughout. Please refer to the TensorFlow dockerfiles documentation
+# for more information.
+
+ARG UBUNTU_VERSION=16.04
+
+FROM ubuntu:${UBUNTU_VERSION} as base
+
+ARG USE_PYTHON_3_NOT_2
+ARG _PY_SUFFIX=${USE_PYTHON_3_NOT_2:+3}
+ARG PYTHON=python${_PY_SUFFIX}
+ARG PIP=pip${_PY_SUFFIX}
+
+# See http://bugs.python.org/issue19846
+ENV LANG C.UTF-8
+
+RUN apt-get update && apt-get install -y \
+    ${PYTHON} \
+    ${PYTHON}-pip
+
+RUN ${PIP} --no-cache-dir install --upgrade \
+    pip \
+    setuptools
+
+# Some TF tools expect a "python" binary
+RUN ln -s $(which ${PYTHON}) /usr/local/bin/python 
+
+# Options:
+#   tensorflow
+#   tensorflow-gpu
+#   tf-nightly
+#   tf-nightly-gpu
+ARG TF_PACKAGE=tensorflow
+RUN apt-get update && apt-get install -y wget libhdf5-dev
+RUN ${PIP} install --global-option=build_ext \
+            --global-option=-I/usr/include/hdf5/serial/ \
+            --global-option=-L/usr/lib/powerpc64le-linux-gnu/hdf5/serial \
+            h5py
+
+# CACHE_STOP is used to rerun future commands, otherwise downloading the .whl will be cached and will not pull the most recent version
+ARG CACHE_STOP=1
+RUN if [ ${TF_PACKAGE} = tensorflow-gpu ]; then \
+        BASE=https://powerci.osuosl.org/job/TensorFlow_PPC64LE_GPU_Release_Build/lastSuccessfulBuild/; \
+    elif [ ${TF_PACKAGE} = tf-nightly-gpu ]; then \
+        BASE=https://powerci.osuosl.org/job/TensorFlow_PPC64LE_GPU_Nightly_Artifact/lastSuccessfulBuild/; \
+    elif [ ${TF_PACKAGE} = tensorflow ]; then \
+        BASE=https://powerci.osuosl.org/job/TensorFlow_PPC64LE_CPU_Release_Build/lastSuccessfulBuild/; \
+    elif [ ${TF_PACKAGE} = tf-nightly ]; then \
+        BASE=https://powerci.osuosl.org/job/TensorFlow_PPC64LE_CPU_Nightly_Artifact/lastSuccessfulBuild/; \
+    fi; \
+    MAJOR=`${PYTHON} -c 'import sys; print(sys.version_info[0])'`; \
+    MINOR=`${PYTHON} -c 'import sys; print(sys.version_info[1])'`; \
+    PACKAGE=$(wget -qO- ${BASE}"api/xml?xpath=//fileName&wrapper=artifacts" | grep -o "[^<>]*cp${MAJOR}${MINOR}[^<>]*.whl"); \
+    wget ${BASE}"artifact/tensorflow_pkg/"${PACKAGE}; \
+    ${PIP} install ${PACKAGE}
+
+COPY bashrc /etc/bash.bashrc
+RUN chmod a+rwx /etc/bash.bashrc
+
+RUN ${PIP} install jupyter matplotlib
+
+RUN mkdir -p /tf/tensorflow-tutorials && chmod -R a+rwx /tf/
+RUN mkdir /.local && chmod a+rwx /.local
+RUN apt-get install -y --no-install-recommends wget
+WORKDIR /tf/tensorflow-tutorials
+RUN wget https://raw.githubusercontent.com/tensorflow/docs/master/site/en/tutorials/keras/basic_classification.ipynb
+RUN wget https://raw.githubusercontent.com/tensorflow/docs/master/site/en/tutorials/keras/basic_text_classification.ipynb
+COPY readme-for-jupyter.md README.md
+RUN apt-get autoremove -y && apt-get remove -y wget
+WORKDIR /tf
+EXPOSE 8888
+
+RUN ${PYTHON} -m ipykernel.kernelspec
+
+CMD ["bash", "-c", "source /etc/bash.bashrc && jupyter notebook --notebook-dir=/tf --ip 0.0.0.0 --no-browser --allow-root"]
diff --git a/tensorflow/tools/dockerfiles/dockerfiles/ppc64le/cpu-ppc64le.Dockerfile b/tensorflow/tools/dockerfiles/dockerfiles/ppc64le/cpu-ppc64le.Dockerfile
new file mode 100644
index 00000000000..083d61bf9a2
--- /dev/null
+++ b/tensorflow/tools/dockerfiles/dockerfiles/ppc64le/cpu-ppc64le.Dockerfile
@@ -0,0 +1,75 @@
+# Copyright 2018 The TensorFlow Authors. All Rights Reserved.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+# ============================================================================
+#
+# THIS IS A GENERATED DOCKERFILE.
+#
+# This file was assembled from multiple pieces, whose use is documented
+# throughout. Please refer to the TensorFlow dockerfiles documentation
+# for more information.
+
+ARG UBUNTU_VERSION=16.04
+
+FROM ubuntu:${UBUNTU_VERSION} as base
+
+ARG USE_PYTHON_3_NOT_2
+ARG _PY_SUFFIX=${USE_PYTHON_3_NOT_2:+3}
+ARG PYTHON=python${_PY_SUFFIX}
+ARG PIP=pip${_PY_SUFFIX}
+
+# See http://bugs.python.org/issue19846
+ENV LANG C.UTF-8
+
+RUN apt-get update && apt-get install -y \
+    ${PYTHON} \
+    ${PYTHON}-pip
+
+RUN ${PIP} --no-cache-dir install --upgrade \
+    pip \
+    setuptools
+
+# Some TF tools expect a "python" binary
+RUN ln -s $(which ${PYTHON}) /usr/local/bin/python 
+
+# Options:
+#   tensorflow
+#   tensorflow-gpu
+#   tf-nightly
+#   tf-nightly-gpu
+ARG TF_PACKAGE=tensorflow
+RUN apt-get update && apt-get install -y wget libhdf5-dev
+RUN ${PIP} install --global-option=build_ext \
+            --global-option=-I/usr/include/hdf5/serial/ \
+            --global-option=-L/usr/lib/powerpc64le-linux-gnu/hdf5/serial \
+            h5py
+
+# CACHE_STOP is used to rerun future commands, otherwise downloading the .whl will be cached and will not pull the most recent version
+ARG CACHE_STOP=1
+RUN if [ ${TF_PACKAGE} = tensorflow-gpu ]; then \
+        BASE=https://powerci.osuosl.org/job/TensorFlow_PPC64LE_GPU_Release_Build/lastSuccessfulBuild/; \
+    elif [ ${TF_PACKAGE} = tf-nightly-gpu ]; then \
+        BASE=https://powerci.osuosl.org/job/TensorFlow_PPC64LE_GPU_Nightly_Artifact/lastSuccessfulBuild/; \
+    elif [ ${TF_PACKAGE} = tensorflow ]; then \
+        BASE=https://powerci.osuosl.org/job/TensorFlow_PPC64LE_CPU_Release_Build/lastSuccessfulBuild/; \
+    elif [ ${TF_PACKAGE} = tf-nightly ]; then \
+        BASE=https://powerci.osuosl.org/job/TensorFlow_PPC64LE_CPU_Nightly_Artifact/lastSuccessfulBuild/; \
+    fi; \
+    MAJOR=`${PYTHON} -c 'import sys; print(sys.version_info[0])'`; \
+    MINOR=`${PYTHON} -c 'import sys; print(sys.version_info[1])'`; \
+    PACKAGE=$(wget -qO- ${BASE}"api/xml?xpath=//fileName&wrapper=artifacts" | grep -o "[^<>]*cp${MAJOR}${MINOR}[^<>]*.whl"); \
+    wget ${BASE}"artifact/tensorflow_pkg/"${PACKAGE}; \
+    ${PIP} install ${PACKAGE}
+
+COPY bashrc /etc/bash.bashrc
+RUN chmod a+rwx /etc/bash.bashrc
diff --git a/tensorflow/tools/dockerfiles/dockerfiles/ppc64le/devel-cpu-ppc64le-jupyter.Dockerfile b/tensorflow/tools/dockerfiles/dockerfiles/ppc64le/devel-cpu-ppc64le-jupyter.Dockerfile
new file mode 100644
index 00000000000..1f32849735c
--- /dev/null
+++ b/tensorflow/tools/dockerfiles/dockerfiles/ppc64le/devel-cpu-ppc64le-jupyter.Dockerfile
@@ -0,0 +1,125 @@
+# Copyright 2018 The TensorFlow Authors. All Rights Reserved.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+# ============================================================================
+#
+# THIS IS A GENERATED DOCKERFILE.
+#
+# This file was assembled from multiple pieces, whose use is documented
+# throughout. Please refer to the TensorFlow dockerfiles documentation
+# for more information.
+
+ARG UBUNTU_VERSION=16.04
+
+FROM ubuntu:${UBUNTU_VERSION} AS base
+
+RUN apt-get update && apt-get install -y --no-install-recommends \
+        build-essential \
+        curl \
+        git \
+        libcurl3-dev \
+        libfreetype6-dev \
+        libhdf5-serial-dev \
+        libzmq3-dev \
+        pkg-config \
+        rsync \
+        software-properties-common \
+        unzip \
+        zip \
+        zlib1g-dev \
+        openjdk-8-jdk \
+        openjdk-8-jre-headless \
+        && \
+    apt-get clean && \
+    rm -rf /var/lib/apt/lists/*
+
+ENV CI_BUILD_PYTHON python
+
+# CACHE_STOP is used to rerun future commands, otherwise cloning tensorflow will be cached and will not pull the most recent version
+ARG CACHE_STOP=1
+# Check out TensorFlow source code if --build_arg CHECKOUT_TF_SRC=1
+ARG CHECKOUT_TF_SRC=0
+RUN test "${CHECKOUT_TF_SRC}" -eq 1 && git clone https://github.com/tensorflow/tensorflow.git /tensorflow_src || true
+
+ARG USE_PYTHON_3_NOT_2
+ARG _PY_SUFFIX=${USE_PYTHON_3_NOT_2:+3}
+ARG PYTHON=python${_PY_SUFFIX}
+ARG PIP=pip${_PY_SUFFIX}
+
+# See http://bugs.python.org/issue19846
+ENV LANG C.UTF-8
+
+RUN apt-get update && apt-get install -y \
+    ${PYTHON} \
+    ${PYTHON}-pip
+
+RUN ${PIP} --no-cache-dir install --upgrade \
+    pip \
+    setuptools
+
+# Some TF tools expect a "python" binary
+RUN ln -s $(which ${PYTHON}) /usr/local/bin/python 
+
+RUN apt-get update && apt-get install -y \
+    build-essential \
+    curl \
+    git \
+    openjdk-8-jdk \
+    ${PYTHON}-dev \
+    swig
+
+RUN ${PIP} --no-cache-dir install \
+    Pillow \
+    h5py \
+    keras_applications \
+    keras_preprocessing \
+    matplotlib \
+    mock \
+    numpy \
+    scipy \
+    sklearn \
+    pandas \
+    && test "${USE_PYTHON_3_NOT_2}" -eq 1 && true || ${PIP} --no-cache-dir install \
+    enum34
+
+ # Build and install bazel
+ENV BAZEL_VERSION 0.15.0
+WORKDIR /
+RUN mkdir /bazel && \
+    cd /bazel && \
+    curl -fSsL -O https://github.com/bazelbuild/bazel/releases/download/$BAZEL_VERSION/bazel-$BAZEL_VERSION-dist.zip && \
+    unzip bazel-$BAZEL_VERSION-dist.zip && \
+    bash ./compile.sh && \
+    cp output/bazel /usr/local/bin/ && \
+    rm -rf /bazel && \
+    cd -
+
+COPY bashrc /etc/bash.bashrc
+RUN chmod a+rwx /etc/bash.bashrc
+
+RUN ${PIP} install jupyter matplotlib
+
+RUN mkdir -p /tf/tensorflow-tutorials && chmod -R a+rwx /tf/
+RUN mkdir /.local && chmod a+rwx /.local
+RUN apt-get install -y --no-install-recommends wget
+WORKDIR /tf/tensorflow-tutorials
+RUN wget https://raw.githubusercontent.com/tensorflow/docs/master/site/en/tutorials/keras/basic_classification.ipynb
+RUN wget https://raw.githubusercontent.com/tensorflow/docs/master/site/en/tutorials/keras/basic_text_classification.ipynb
+COPY readme-for-jupyter.md README.md
+RUN apt-get autoremove -y && apt-get remove -y wget
+WORKDIR /tf
+EXPOSE 8888
+
+RUN ${PYTHON} -m ipykernel.kernelspec
+
+CMD ["bash", "-c", "source /etc/bash.bashrc && jupyter notebook --notebook-dir=/tf --ip 0.0.0.0 --no-browser --allow-root"]
diff --git a/tensorflow/tools/dockerfiles/dockerfiles/ppc64le/devel-cpu-ppc64le.Dockerfile b/tensorflow/tools/dockerfiles/dockerfiles/ppc64le/devel-cpu-ppc64le.Dockerfile
new file mode 100644
index 00000000000..cda51c371d0
--- /dev/null
+++ b/tensorflow/tools/dockerfiles/dockerfiles/ppc64le/devel-cpu-ppc64le.Dockerfile
@@ -0,0 +1,108 @@
+# Copyright 2018 The TensorFlow Authors. All Rights Reserved.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+# ============================================================================
+#
+# THIS IS A GENERATED DOCKERFILE.
+#
+# This file was assembled from multiple pieces, whose use is documented
+# throughout. Please refer to the TensorFlow dockerfiles documentation
+# for more information.
+
+ARG UBUNTU_VERSION=16.04
+
+FROM ubuntu:${UBUNTU_VERSION} AS base
+
+RUN apt-get update && apt-get install -y --no-install-recommends \
+        build-essential \
+        curl \
+        git \
+        libcurl3-dev \
+        libfreetype6-dev \
+        libhdf5-serial-dev \
+        libzmq3-dev \
+        pkg-config \
+        rsync \
+        software-properties-common \
+        unzip \
+        zip \
+        zlib1g-dev \
+        openjdk-8-jdk \
+        openjdk-8-jre-headless \
+        && \
+    apt-get clean && \
+    rm -rf /var/lib/apt/lists/*
+
+ENV CI_BUILD_PYTHON python
+
+# CACHE_STOP is used to rerun future commands, otherwise cloning tensorflow will be cached and will not pull the most recent version
+ARG CACHE_STOP=1
+# Check out TensorFlow source code if --build_arg CHECKOUT_TF_SRC=1
+ARG CHECKOUT_TF_SRC=0
+RUN test "${CHECKOUT_TF_SRC}" -eq 1 && git clone https://github.com/tensorflow/tensorflow.git /tensorflow_src || true
+
+ARG USE_PYTHON_3_NOT_2
+ARG _PY_SUFFIX=${USE_PYTHON_3_NOT_2:+3}
+ARG PYTHON=python${_PY_SUFFIX}
+ARG PIP=pip${_PY_SUFFIX}
+
+# See http://bugs.python.org/issue19846
+ENV LANG C.UTF-8
+
+RUN apt-get update && apt-get install -y \
+    ${PYTHON} \
+    ${PYTHON}-pip
+
+RUN ${PIP} --no-cache-dir install --upgrade \
+    pip \
+    setuptools
+
+# Some TF tools expect a "python" binary
+RUN ln -s $(which ${PYTHON}) /usr/local/bin/python 
+
+RUN apt-get update && apt-get install -y \
+    build-essential \
+    curl \
+    git \
+    openjdk-8-jdk \
+    ${PYTHON}-dev \
+    swig
+
+RUN ${PIP} --no-cache-dir install \
+    Pillow \
+    h5py \
+    keras_applications \
+    keras_preprocessing \
+    matplotlib \
+    mock \
+    numpy \
+    scipy \
+    sklearn \
+    pandas \
+    && test "${USE_PYTHON_3_NOT_2}" -eq 1 && true || ${PIP} --no-cache-dir install \
+    enum34
+
+ # Build and install bazel
+ENV BAZEL_VERSION 0.15.0
+WORKDIR /
+RUN mkdir /bazel && \
+    cd /bazel && \
+    curl -fSsL -O https://github.com/bazelbuild/bazel/releases/download/$BAZEL_VERSION/bazel-$BAZEL_VERSION-dist.zip && \
+    unzip bazel-$BAZEL_VERSION-dist.zip && \
+    bash ./compile.sh && \
+    cp output/bazel /usr/local/bin/ && \
+    rm -rf /bazel && \
+    cd -
+
+COPY bashrc /etc/bash.bashrc
+RUN chmod a+rwx /etc/bash.bashrc
diff --git a/tensorflow/tools/dockerfiles/dockerfiles/ppc64le/devel-gpu-ppc64le-jupyter.Dockerfile b/tensorflow/tools/dockerfiles/dockerfiles/ppc64le/devel-gpu-ppc64le-jupyter.Dockerfile
new file mode 100644
index 00000000000..d8ee19f66ef
--- /dev/null
+++ b/tensorflow/tools/dockerfiles/dockerfiles/ppc64le/devel-gpu-ppc64le-jupyter.Dockerfile
@@ -0,0 +1,151 @@
+# Copyright 2018 The TensorFlow Authors. All Rights Reserved.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+# ============================================================================
+#
+# THIS IS A GENERATED DOCKERFILE.
+#
+# This file was assembled from multiple pieces, whose use is documented
+# throughout. Please refer to the TensorFlow dockerfiles documentation
+# for more information.
+
+ARG UBUNTU_VERSION=16.04
+
+ARG ARCH=
+ARG CUDA=10.0
+FROM nvidia/cuda${ARCH:+-$ARCH}:${CUDA}-base-ubuntu${UBUNTU_VERSION} as base
+ARG CUDNN=7.4.1.5-1
+ARG LIB_DIR_PREFIX=x84_64
+
+# Needed for string substitution 
+SHELL ["/bin/bash", "-c"]
+RUN apt-get update && apt-get install -y --no-install-recommends \
+        build-essential \
+        cuda-command-line-tools-${CUDA/./-} \
+        cuda-cublas-dev-${CUDA/./-} \
+        cuda-cudart-dev-${CUDA/./-} \
+        cuda-cufft-dev-${CUDA/./-} \
+        cuda-curand-dev-${CUDA/./-} \
+        cuda-cusolver-dev-${CUDA/./-} \
+        cuda-cusparse-dev-${CUDA/./-} \
+        libcudnn7=${CUDNN}+cuda${CUDA} \
+        libcudnn7-dev=${CUDNN}+cuda${CUDA} \
+        libcurl3-dev \
+        libfreetype6-dev \
+        libhdf5-serial-dev \
+        libzmq3-dev \
+        pkg-config \
+        rsync \
+        software-properties-common \
+        unzip \
+        zip \
+        zlib1g-dev \
+        wget \
+        git \
+        && \
+    find /usr/local/cuda-${CUDA}/lib64/ -type f -name 'lib*_static.a' -not -name 'libcudart_static.a' -delete && \
+    rm /usr/lib/${LIB_DIR_PREFIX}-linux-gnu/libcudnn_static_v7.a
+
+RUN [ ${ARCH} = ppc64le ] || (apt-get update && \
+        apt-get install nvinfer-runtime-trt-repo-ubuntu1604-5.0.2-ga-cuda${CUDA} \
+        && apt-get update \
+        && apt-get install -y --no-install-recommends libnvinfer5=5.0.2-1+cuda${CUDA} \
+        && apt-get clean \
+        && rm -rf /var/lib/apt/lists/*)
+
+# Configure the build for our CUDA configuration.
+ENV CI_BUILD_PYTHON python
+ENV LD_LIBRARY_PATH /usr/local/cuda/extras/CUPTI/lib64:$LD_LIBRARY_PATH
+ENV TF_NEED_CUDA 1
+ENV TF_NEED_TENSORRT 1
+ENV TF_CUDA_COMPUTE_CAPABILITIES=3.5,5.2,6.0,6.1,7.0
+ENV TF_CUDA_VERSION=${CUDA}
+ENV TF_CUDNN_VERSION=${CUDNN%%.*}
+# CACHE_STOP is used to rerun future commands, otherwise cloning tensorflow will be cached and will not pull the most recent version
+ARG CACHE_STOP=1
+# Check out TensorFlow source code if --build_arg CHECKOUT_TF_SRC=1
+ARG CHECKOUT_TF_SRC=0
+RUN test "${CHECKOUT_TF_SRC}" -eq 1 && git clone https://github.com/tensorflow/tensorflow.git /tensorflow_src || true
+
+ARG USE_PYTHON_3_NOT_2
+ARG _PY_SUFFIX=${USE_PYTHON_3_NOT_2:+3}
+ARG PYTHON=python${_PY_SUFFIX}
+ARG PIP=pip${_PY_SUFFIX}
+
+# See http://bugs.python.org/issue19846
+ENV LANG C.UTF-8
+
+RUN apt-get update && apt-get install -y \
+    ${PYTHON} \
+    ${PYTHON}-pip
+
+RUN ${PIP} --no-cache-dir install --upgrade \
+    pip \
+    setuptools
+
+# Some TF tools expect a "python" binary
+RUN ln -s $(which ${PYTHON}) /usr/local/bin/python 
+
+RUN apt-get update && apt-get install -y \
+    build-essential \
+    curl \
+    git \
+    openjdk-8-jdk \
+    ${PYTHON}-dev \
+    swig
+
+RUN ${PIP} --no-cache-dir install \
+    Pillow \
+    h5py \
+    keras_applications \
+    keras_preprocessing \
+    matplotlib \
+    mock \
+    numpy \
+    scipy \
+    sklearn \
+    pandas \
+    && test "${USE_PYTHON_3_NOT_2}" -eq 1 && true || ${PIP} --no-cache-dir install \
+    enum34
+
+ # Build and install bazel
+ENV BAZEL_VERSION 0.15.0
+WORKDIR /
+RUN mkdir /bazel && \
+    cd /bazel && \
+    curl -fSsL -O https://github.com/bazelbuild/bazel/releases/download/$BAZEL_VERSION/bazel-$BAZEL_VERSION-dist.zip && \
+    unzip bazel-$BAZEL_VERSION-dist.zip && \
+    bash ./compile.sh && \
+    cp output/bazel /usr/local/bin/ && \
+    rm -rf /bazel && \
+    cd -
+
+COPY bashrc /etc/bash.bashrc
+RUN chmod a+rwx /etc/bash.bashrc
+
+RUN ${PIP} install jupyter matplotlib
+
+RUN mkdir -p /tf/tensorflow-tutorials && chmod -R a+rwx /tf/
+RUN mkdir /.local && chmod a+rwx /.local
+RUN apt-get install -y --no-install-recommends wget
+WORKDIR /tf/tensorflow-tutorials
+RUN wget https://raw.githubusercontent.com/tensorflow/docs/master/site/en/tutorials/keras/basic_classification.ipynb
+RUN wget https://raw.githubusercontent.com/tensorflow/docs/master/site/en/tutorials/keras/basic_text_classification.ipynb
+COPY readme-for-jupyter.md README.md
+RUN apt-get autoremove -y && apt-get remove -y wget
+WORKDIR /tf
+EXPOSE 8888
+
+RUN ${PYTHON} -m ipykernel.kernelspec
+
+CMD ["bash", "-c", "source /etc/bash.bashrc && jupyter notebook --notebook-dir=/tf --ip 0.0.0.0 --no-browser --allow-root"]
diff --git a/tensorflow/tools/dockerfiles/dockerfiles/ppc64le/devel-gpu-ppc64le.Dockerfile b/tensorflow/tools/dockerfiles/dockerfiles/ppc64le/devel-gpu-ppc64le.Dockerfile
new file mode 100644
index 00000000000..966070634b8
--- /dev/null
+++ b/tensorflow/tools/dockerfiles/dockerfiles/ppc64le/devel-gpu-ppc64le.Dockerfile
@@ -0,0 +1,134 @@
+# Copyright 2018 The TensorFlow Authors. All Rights Reserved.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+# ============================================================================
+#
+# THIS IS A GENERATED DOCKERFILE.
+#
+# This file was assembled from multiple pieces, whose use is documented
+# throughout. Please refer to the TensorFlow dockerfiles documentation
+# for more information.
+
+ARG UBUNTU_VERSION=16.04
+
+ARG ARCH=
+ARG CUDA=10.0
+FROM nvidia/cuda${ARCH:+-$ARCH}:${CUDA}-base-ubuntu${UBUNTU_VERSION} as base
+ARG CUDNN=7.4.1.5-1
+ARG LIB_DIR_PREFIX=x84_64
+
+# Needed for string substitution 
+SHELL ["/bin/bash", "-c"]
+RUN apt-get update && apt-get install -y --no-install-recommends \
+        build-essential \
+        cuda-command-line-tools-${CUDA/./-} \
+        cuda-cublas-dev-${CUDA/./-} \
+        cuda-cudart-dev-${CUDA/./-} \
+        cuda-cufft-dev-${CUDA/./-} \
+        cuda-curand-dev-${CUDA/./-} \
+        cuda-cusolver-dev-${CUDA/./-} \
+        cuda-cusparse-dev-${CUDA/./-} \
+        libcudnn7=${CUDNN}+cuda${CUDA} \
+        libcudnn7-dev=${CUDNN}+cuda${CUDA} \
+        libcurl3-dev \
+        libfreetype6-dev \
+        libhdf5-serial-dev \
+        libzmq3-dev \
+        pkg-config \
+        rsync \
+        software-properties-common \
+        unzip \
+        zip \
+        zlib1g-dev \
+        wget \
+        git \
+        && \
+    find /usr/local/cuda-${CUDA}/lib64/ -type f -name 'lib*_static.a' -not -name 'libcudart_static.a' -delete && \
+    rm /usr/lib/${LIB_DIR_PREFIX}-linux-gnu/libcudnn_static_v7.a
+
+RUN [ ${ARCH} = ppc64le ] || (apt-get update && \
+        apt-get install nvinfer-runtime-trt-repo-ubuntu1604-5.0.2-ga-cuda${CUDA} \
+        && apt-get update \
+        && apt-get install -y --no-install-recommends libnvinfer5=5.0.2-1+cuda${CUDA} \
+        && apt-get clean \
+        && rm -rf /var/lib/apt/lists/*)
+
+# Configure the build for our CUDA configuration.
+ENV CI_BUILD_PYTHON python
+ENV LD_LIBRARY_PATH /usr/local/cuda/extras/CUPTI/lib64:$LD_LIBRARY_PATH
+ENV TF_NEED_CUDA 1
+ENV TF_NEED_TENSORRT 1
+ENV TF_CUDA_COMPUTE_CAPABILITIES=3.5,5.2,6.0,6.1,7.0
+ENV TF_CUDA_VERSION=${CUDA}
+ENV TF_CUDNN_VERSION=${CUDNN%%.*}
+# CACHE_STOP is used to rerun future commands, otherwise cloning tensorflow will be cached and will not pull the most recent version
+ARG CACHE_STOP=1
+# Check out TensorFlow source code if --build_arg CHECKOUT_TF_SRC=1
+ARG CHECKOUT_TF_SRC=0
+RUN test "${CHECKOUT_TF_SRC}" -eq 1 && git clone https://github.com/tensorflow/tensorflow.git /tensorflow_src || true
+
+ARG USE_PYTHON_3_NOT_2
+ARG _PY_SUFFIX=${USE_PYTHON_3_NOT_2:+3}
+ARG PYTHON=python${_PY_SUFFIX}
+ARG PIP=pip${_PY_SUFFIX}
+
+# See http://bugs.python.org/issue19846
+ENV LANG C.UTF-8
+
+RUN apt-get update && apt-get install -y \
+    ${PYTHON} \
+    ${PYTHON}-pip
+
+RUN ${PIP} --no-cache-dir install --upgrade \
+    pip \
+    setuptools
+
+# Some TF tools expect a "python" binary
+RUN ln -s $(which ${PYTHON}) /usr/local/bin/python 
+
+RUN apt-get update && apt-get install -y \
+    build-essential \
+    curl \
+    git \
+    openjdk-8-jdk \
+    ${PYTHON}-dev \
+    swig
+
+RUN ${PIP} --no-cache-dir install \
+    Pillow \
+    h5py \
+    keras_applications \
+    keras_preprocessing \
+    matplotlib \
+    mock \
+    numpy \
+    scipy \
+    sklearn \
+    pandas \
+    && test "${USE_PYTHON_3_NOT_2}" -eq 1 && true || ${PIP} --no-cache-dir install \
+    enum34
+
+ # Build and install bazel
+ENV BAZEL_VERSION 0.15.0
+WORKDIR /
+RUN mkdir /bazel && \
+    cd /bazel && \
+    curl -fSsL -O https://github.com/bazelbuild/bazel/releases/download/$BAZEL_VERSION/bazel-$BAZEL_VERSION-dist.zip && \
+    unzip bazel-$BAZEL_VERSION-dist.zip && \
+    bash ./compile.sh && \
+    cp output/bazel /usr/local/bin/ && \
+    rm -rf /bazel && \
+    cd -
+
+COPY bashrc /etc/bash.bashrc
+RUN chmod a+rwx /etc/bash.bashrc
diff --git a/tensorflow/tools/dockerfiles/dockerfiles/ppc64le/gpu-ppc64le-jupyter.Dockerfile b/tensorflow/tools/dockerfiles/dockerfiles/ppc64le/gpu-ppc64le-jupyter.Dockerfile
new file mode 100644
index 00000000000..449a8d8aa89
--- /dev/null
+++ b/tensorflow/tools/dockerfiles/dockerfiles/ppc64le/gpu-ppc64le-jupyter.Dockerfile
@@ -0,0 +1,125 @@
+# Copyright 2018 The TensorFlow Authors. All Rights Reserved.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+# ============================================================================
+#
+# THIS IS A GENERATED DOCKERFILE.
+#
+# This file was assembled from multiple pieces, whose use is documented
+# throughout. Please refer to the TensorFlow dockerfiles documentation
+# for more information.
+
+ARG UBUNTU_VERSION=16.04
+
+ARG ARCH=
+ARG CUDA=10.0
+FROM nvidia/cuda${ARCH:+-$ARCH}:${CUDA}-base-ubuntu${UBUNTU_VERSION} as base
+ARG CUDNN=7.4.1.5-1
+
+# Needed for string substitution 
+SHELL ["/bin/bash", "-c"]
+# Pick up some TF dependencies
+RUN apt-get update && apt-get install -y --no-install-recommends \
+        build-essential \
+        cuda-command-line-tools-${CUDA/./-} \
+        cuda-cublas-${CUDA/./-} \
+        cuda-cufft-${CUDA/./-} \
+        cuda-curand-${CUDA/./-} \
+        cuda-cusolver-${CUDA/./-} \
+        cuda-cusparse-${CUDA/./-} \
+        curl \
+        libcudnn7=${CUDNN}+cuda${CUDA} \
+        libfreetype6-dev \
+        libhdf5-serial-dev \
+        libzmq3-dev \
+        pkg-config \
+        software-properties-common \
+        unzip
+
+RUN [ ${ARCH} = ppc64le ] || (apt-get update && \
+        apt-get install nvinfer-runtime-trt-repo-ubuntu1604-5.0.2-ga-cuda${CUDA} \
+        && apt-get update \
+        && apt-get install -y --no-install-recommends libnvinfer5=5.0.2-1+cuda${CUDA} \
+        && apt-get clean \
+        && rm -rf /var/lib/apt/lists/*)
+
+# For CUDA profiling, TensorFlow requires CUPTI.
+ENV LD_LIBRARY_PATH /usr/local/cuda/extras/CUPTI/lib64:$LD_LIBRARY_PATH
+
+ARG USE_PYTHON_3_NOT_2
+ARG _PY_SUFFIX=${USE_PYTHON_3_NOT_2:+3}
+ARG PYTHON=python${_PY_SUFFIX}
+ARG PIP=pip${_PY_SUFFIX}
+
+# See http://bugs.python.org/issue19846
+ENV LANG C.UTF-8
+
+RUN apt-get update && apt-get install -y \
+    ${PYTHON} \
+    ${PYTHON}-pip
+
+RUN ${PIP} --no-cache-dir install --upgrade \
+    pip \
+    setuptools
+
+# Some TF tools expect a "python" binary
+RUN ln -s $(which ${PYTHON}) /usr/local/bin/python 
+
+# Options:
+#   tensorflow
+#   tensorflow-gpu
+#   tf-nightly
+#   tf-nightly-gpu
+ARG TF_PACKAGE=tensorflow
+RUN apt-get update && apt-get install -y wget libhdf5-dev
+RUN ${PIP} install --global-option=build_ext \
+            --global-option=-I/usr/include/hdf5/serial/ \
+            --global-option=-L/usr/lib/powerpc64le-linux-gnu/hdf5/serial \
+            h5py
+
+# CACHE_STOP is used to rerun future commands, otherwise downloading the .whl will be cached and will not pull the most recent version
+ARG CACHE_STOP=1
+RUN if [ ${TF_PACKAGE} = tensorflow-gpu ]; then \
+        BASE=https://powerci.osuosl.org/job/TensorFlow_PPC64LE_GPU_Release_Build/lastSuccessfulBuild/; \
+    elif [ ${TF_PACKAGE} = tf-nightly-gpu ]; then \
+        BASE=https://powerci.osuosl.org/job/TensorFlow_PPC64LE_GPU_Nightly_Artifact/lastSuccessfulBuild/; \
+    elif [ ${TF_PACKAGE} = tensorflow ]; then \
+        BASE=https://powerci.osuosl.org/job/TensorFlow_PPC64LE_CPU_Release_Build/lastSuccessfulBuild/; \
+    elif [ ${TF_PACKAGE} = tf-nightly ]; then \
+        BASE=https://powerci.osuosl.org/job/TensorFlow_PPC64LE_CPU_Nightly_Artifact/lastSuccessfulBuild/; \
+    fi; \
+    MAJOR=`${PYTHON} -c 'import sys; print(sys.version_info[0])'`; \
+    MINOR=`${PYTHON} -c 'import sys; print(sys.version_info[1])'`; \
+    PACKAGE=$(wget -qO- ${BASE}"api/xml?xpath=//fileName&wrapper=artifacts" | grep -o "[^<>]*cp${MAJOR}${MINOR}[^<>]*.whl"); \
+    wget ${BASE}"artifact/tensorflow_pkg/"${PACKAGE}; \
+    ${PIP} install ${PACKAGE}
+
+COPY bashrc /etc/bash.bashrc
+RUN chmod a+rwx /etc/bash.bashrc
+
+RUN ${PIP} install jupyter matplotlib
+
+RUN mkdir -p /tf/tensorflow-tutorials && chmod -R a+rwx /tf/
+RUN mkdir /.local && chmod a+rwx /.local
+RUN apt-get install -y --no-install-recommends wget
+WORKDIR /tf/tensorflow-tutorials
+RUN wget https://raw.githubusercontent.com/tensorflow/docs/master/site/en/tutorials/keras/basic_classification.ipynb
+RUN wget https://raw.githubusercontent.com/tensorflow/docs/master/site/en/tutorials/keras/basic_text_classification.ipynb
+COPY readme-for-jupyter.md README.md
+RUN apt-get autoremove -y && apt-get remove -y wget
+WORKDIR /tf
+EXPOSE 8888
+
+RUN ${PYTHON} -m ipykernel.kernelspec
+
+CMD ["bash", "-c", "source /etc/bash.bashrc && jupyter notebook --notebook-dir=/tf --ip 0.0.0.0 --no-browser --allow-root"]
diff --git a/tensorflow/tools/dockerfiles/dockerfiles/ppc64le/gpu-ppc64le.Dockerfile b/tensorflow/tools/dockerfiles/dockerfiles/ppc64le/gpu-ppc64le.Dockerfile
new file mode 100644
index 00000000000..f01a47f1c02
--- /dev/null
+++ b/tensorflow/tools/dockerfiles/dockerfiles/ppc64le/gpu-ppc64le.Dockerfile
@@ -0,0 +1,108 @@
+# Copyright 2018 The TensorFlow Authors. All Rights Reserved.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+# ============================================================================
+#
+# THIS IS A GENERATED DOCKERFILE.
+#
+# This file was assembled from multiple pieces, whose use is documented
+# throughout. Please refer to the TensorFlow dockerfiles documentation
+# for more information.
+
+ARG UBUNTU_VERSION=16.04
+
+ARG ARCH=
+ARG CUDA=10.0
+FROM nvidia/cuda${ARCH:+-$ARCH}:${CUDA}-base-ubuntu${UBUNTU_VERSION} as base
+ARG CUDNN=7.4.1.5-1
+
+# Needed for string substitution 
+SHELL ["/bin/bash", "-c"]
+# Pick up some TF dependencies
+RUN apt-get update && apt-get install -y --no-install-recommends \
+        build-essential \
+        cuda-command-line-tools-${CUDA/./-} \
+        cuda-cublas-${CUDA/./-} \
+        cuda-cufft-${CUDA/./-} \
+        cuda-curand-${CUDA/./-} \
+        cuda-cusolver-${CUDA/./-} \
+        cuda-cusparse-${CUDA/./-} \
+        curl \
+        libcudnn7=${CUDNN}+cuda${CUDA} \
+        libfreetype6-dev \
+        libhdf5-serial-dev \
+        libzmq3-dev \
+        pkg-config \
+        software-properties-common \
+        unzip
+
+RUN [ ${ARCH} = ppc64le ] || (apt-get update && \
+        apt-get install nvinfer-runtime-trt-repo-ubuntu1604-5.0.2-ga-cuda${CUDA} \
+        && apt-get update \
+        && apt-get install -y --no-install-recommends libnvinfer5=5.0.2-1+cuda${CUDA} \
+        && apt-get clean \
+        && rm -rf /var/lib/apt/lists/*)
+
+# For CUDA profiling, TensorFlow requires CUPTI.
+ENV LD_LIBRARY_PATH /usr/local/cuda/extras/CUPTI/lib64:$LD_LIBRARY_PATH
+
+ARG USE_PYTHON_3_NOT_2
+ARG _PY_SUFFIX=${USE_PYTHON_3_NOT_2:+3}
+ARG PYTHON=python${_PY_SUFFIX}
+ARG PIP=pip${_PY_SUFFIX}
+
+# See http://bugs.python.org/issue19846
+ENV LANG C.UTF-8
+
+RUN apt-get update && apt-get install -y \
+    ${PYTHON} \
+    ${PYTHON}-pip
+
+RUN ${PIP} --no-cache-dir install --upgrade \
+    pip \
+    setuptools
+
+# Some TF tools expect a "python" binary
+RUN ln -s $(which ${PYTHON}) /usr/local/bin/python 
+
+# Options:
+#   tensorflow
+#   tensorflow-gpu
+#   tf-nightly
+#   tf-nightly-gpu
+ARG TF_PACKAGE=tensorflow
+RUN apt-get update && apt-get install -y wget libhdf5-dev
+RUN ${PIP} install --global-option=build_ext \
+            --global-option=-I/usr/include/hdf5/serial/ \
+            --global-option=-L/usr/lib/powerpc64le-linux-gnu/hdf5/serial \
+            h5py
+
+# CACHE_STOP is used to rerun future commands, otherwise downloading the .whl will be cached and will not pull the most recent version
+ARG CACHE_STOP=1
+RUN if [ ${TF_PACKAGE} = tensorflow-gpu ]; then \
+        BASE=https://powerci.osuosl.org/job/TensorFlow_PPC64LE_GPU_Release_Build/lastSuccessfulBuild/; \
+    elif [ ${TF_PACKAGE} = tf-nightly-gpu ]; then \
+        BASE=https://powerci.osuosl.org/job/TensorFlow_PPC64LE_GPU_Nightly_Artifact/lastSuccessfulBuild/; \
+    elif [ ${TF_PACKAGE} = tensorflow ]; then \
+        BASE=https://powerci.osuosl.org/job/TensorFlow_PPC64LE_CPU_Release_Build/lastSuccessfulBuild/; \
+    elif [ ${TF_PACKAGE} = tf-nightly ]; then \
+        BASE=https://powerci.osuosl.org/job/TensorFlow_PPC64LE_CPU_Nightly_Artifact/lastSuccessfulBuild/; \
+    fi; \
+    MAJOR=`${PYTHON} -c 'import sys; print(sys.version_info[0])'`; \
+    MINOR=`${PYTHON} -c 'import sys; print(sys.version_info[1])'`; \
+    PACKAGE=$(wget -qO- ${BASE}"api/xml?xpath=//fileName&wrapper=artifacts" | grep -o "[^<>]*cp${MAJOR}${MINOR}[^<>]*.whl"); \
+    wget ${BASE}"artifact/tensorflow_pkg/"${PACKAGE}; \
+    ${PIP} install ${PACKAGE}
+
+COPY bashrc /etc/bash.bashrc
+RUN chmod a+rwx /etc/bash.bashrc
diff --git a/tensorflow/tools/dockerfiles/partials/tensorflow-ppc64le.partial.Dockerfile b/tensorflow/tools/dockerfiles/partials/tensorflow-ppc64le.partial.Dockerfile
new file mode 100644
index 00000000000..1e79574a34d
--- /dev/null
+++ b/tensorflow/tools/dockerfiles/partials/tensorflow-ppc64le.partial.Dockerfile
@@ -0,0 +1,28 @@
+# Options:
+#   tensorflow
+#   tensorflow-gpu
+#   tf-nightly
+#   tf-nightly-gpu
+ARG TF_PACKAGE=tensorflow
+RUN apt-get update && apt-get install -y wget libhdf5-dev
+RUN ${PIP} install --global-option=build_ext \
+            --global-option=-I/usr/include/hdf5/serial/ \
+            --global-option=-L/usr/lib/powerpc64le-linux-gnu/hdf5/serial \
+            h5py
+
+# CACHE_STOP is used to rerun future commands, otherwise downloading the .whl will be cached and will not pull the most recent version
+ARG CACHE_STOP=1
+RUN if [ ${TF_PACKAGE} = tensorflow-gpu ]; then \
+        BASE=https://powerci.osuosl.org/job/TensorFlow_PPC64LE_GPU_Release_Build/lastSuccessfulBuild/; \
+    elif [ ${TF_PACKAGE} = tf-nightly-gpu ]; then \
+        BASE=https://powerci.osuosl.org/job/TensorFlow_PPC64LE_GPU_Nightly_Artifact/lastSuccessfulBuild/; \
+    elif [ ${TF_PACKAGE} = tensorflow ]; then \
+        BASE=https://powerci.osuosl.org/job/TensorFlow_PPC64LE_CPU_Release_Build/lastSuccessfulBuild/; \
+    elif [ ${TF_PACKAGE} = tf-nightly ]; then \
+        BASE=https://powerci.osuosl.org/job/TensorFlow_PPC64LE_CPU_Nightly_Artifact/lastSuccessfulBuild/; \
+    fi; \
+    MAJOR=`${PYTHON} -c 'import sys; print(sys.version_info[0])'`; \
+    MINOR=`${PYTHON} -c 'import sys; print(sys.version_info[1])'`; \
+    PACKAGE=$(wget -qO- ${BASE}"api/xml?xpath=//fileName&wrapper=artifacts" | grep -o "[^<>]*cp${MAJOR}${MINOR}[^<>]*.whl"); \
+    wget ${BASE}"artifact/tensorflow_pkg/"${PACKAGE}; \
+    ${PIP} install ${PACKAGE}
diff --git a/tensorflow/tools/dockerfiles/partials/ubuntu/bazelbuild.partial.Dockerfile b/tensorflow/tools/dockerfiles/partials/ubuntu/bazelbuild.partial.Dockerfile
new file mode 100644
index 00000000000..0397ab5fa85
--- /dev/null
+++ b/tensorflow/tools/dockerfiles/partials/ubuntu/bazelbuild.partial.Dockerfile
@@ -0,0 +1,33 @@
+RUN apt-get update && apt-get install -y \
+    build-essential \
+    curl \
+    git \
+    openjdk-8-jdk \
+    ${PYTHON}-dev \
+    swig
+
+RUN ${PIP} --no-cache-dir install \
+    Pillow \
+    h5py \
+    keras_applications \
+    keras_preprocessing \
+    matplotlib \
+    mock \
+    numpy \
+    scipy \
+    sklearn \
+    pandas \
+    && test "${USE_PYTHON_3_NOT_2}" -eq 1 && true || ${PIP} --no-cache-dir install \
+    enum34
+
+ # Build and install bazel
+ENV BAZEL_VERSION 0.15.0
+WORKDIR /
+RUN mkdir /bazel && \
+    cd /bazel && \
+    curl -fSsL -O https://github.com/bazelbuild/bazel/releases/download/$BAZEL_VERSION/bazel-$BAZEL_VERSION-dist.zip && \
+    unzip bazel-$BAZEL_VERSION-dist.zip && \
+    bash ./compile.sh && \
+    cp output/bazel /usr/local/bin/ && \
+    rm -rf /bazel && \
+    cd -
diff --git a/tensorflow/tools/dockerfiles/partials/ubuntu/devel-cpu.partial.Dockerfile b/tensorflow/tools/dockerfiles/partials/ubuntu/devel-cpu.partial.Dockerfile
index 0652ac4151d..aaeda0b2077 100644
--- a/tensorflow/tools/dockerfiles/partials/ubuntu/devel-cpu.partial.Dockerfile
+++ b/tensorflow/tools/dockerfiles/partials/ubuntu/devel-cpu.partial.Dockerfile
@@ -7,7 +7,6 @@ RUN apt-get update && apt-get install -y --no-install-recommends \
         libcurl3-dev \
         libfreetype6-dev \
         libhdf5-serial-dev \
-        libpng12-dev \
         libzmq3-dev \
         pkg-config \
         rsync \
@@ -20,9 +19,11 @@ RUN apt-get update && apt-get install -y --no-install-recommends \
         && \
     apt-get clean && \
     rm -rf /var/lib/apt/lists/*
- 
+
 ENV CI_BUILD_PYTHON python
 
-# Check out TensorFlow source code if --build-arg CHECKOUT_TF_SRC=1
+# CACHE_STOP is used to rerun future commands, otherwise cloning tensorflow will be cached and will not pull the most recent version
+ARG CACHE_STOP=1
+# Check out TensorFlow source code if --build_arg CHECKOUT_TF_SRC=1
 ARG CHECKOUT_TF_SRC=0
-RUN test "${CHECKOUT_TF_SRC}" -eq 1 && git clone https://github.com/tensorflow/tensorflow.git /tensorflow_src
+RUN test "${CHECKOUT_TF_SRC}" -eq 1 && git clone https://github.com/tensorflow/tensorflow.git /tensorflow_src || true
diff --git a/tensorflow/tools/dockerfiles/partials/ubuntu/devel-nvidia.partial.Dockerfile b/tensorflow/tools/dockerfiles/partials/ubuntu/devel-nvidia.partial.Dockerfile
index 2b4494ac595..8ce4a1879d1 100644
--- a/tensorflow/tools/dockerfiles/partials/ubuntu/devel-nvidia.partial.Dockerfile
+++ b/tensorflow/tools/dockerfiles/partials/ubuntu/devel-nvidia.partial.Dockerfile
@@ -1,20 +1,25 @@
-FROM nvidia/cuda:10.0-base-ubuntu${UBUNTU_VERSION} as base
+ARG ARCH=
+ARG CUDA=10.0
+FROM nvidia/cuda${ARCH:+-$ARCH}:${CUDA}-base-ubuntu${UBUNTU_VERSION} as base
+ARG CUDNN=7.4.1.5-1
+ARG LIB_DIR_PREFIX=x84_64
 
+# Needed for string substitution 
+SHELL ["/bin/bash", "-c"]
 RUN apt-get update && apt-get install -y --no-install-recommends \
         build-essential \
-        cuda-command-line-tools-10-0 \
-        cuda-cublas-dev-10-0 \
-        cuda-cudart-dev-10-0 \
-        cuda-cufft-dev-10-0 \
-        cuda-curand-dev-10-0 \
-        cuda-cusolver-dev-10-0 \
-        cuda-cusparse-dev-10-0 \
-        libcudnn7=7.4.1.5-1+cuda10.0 \
-        libcudnn7-dev=7.4.1.5-1+cuda10.0 \
+        cuda-command-line-tools-${CUDA/./-} \
+        cuda-cublas-dev-${CUDA/./-} \
+        cuda-cudart-dev-${CUDA/./-} \
+        cuda-cufft-dev-${CUDA/./-} \
+        cuda-curand-dev-${CUDA/./-} \
+        cuda-cusolver-dev-${CUDA/./-} \
+        cuda-cusparse-dev-${CUDA/./-} \
+        libcudnn7=${CUDNN}+cuda${CUDA} \
+        libcudnn7-dev=${CUDNN}+cuda${CUDA} \
         libcurl3-dev \
         libfreetype6-dev \
         libhdf5-serial-dev \
-        libpng12-dev \
         libzmq3-dev \
         pkg-config \
         rsync \
@@ -25,14 +30,15 @@ RUN apt-get update && apt-get install -y --no-install-recommends \
         wget \
         git \
         && \
-    find /usr/local/cuda-10.0/lib64/ -type f -name 'lib*_static.a' -not -name 'libcudart_static.a' -delete && \
-    rm /usr/lib/x86_64-linux-gnu/libcudnn_static_v7.a
+    find /usr/local/cuda-${CUDA}/lib64/ -type f -name 'lib*_static.a' -not -name 'libcudart_static.a' -delete && \
+    rm /usr/lib/${LIB_DIR_PREFIX}-linux-gnu/libcudnn_static_v7.a
 
-RUN apt-get update && \
-        apt-get install nvinfer-runtime-trt-repo-ubuntu1604-5.0.2-ga-cuda10.0 \
+RUN [ ${ARCH} = ppc64le ] || (apt-get update && \
+        apt-get install nvinfer-runtime-trt-repo-ubuntu1604-5.0.2-ga-cuda${CUDA} \
         && apt-get update \
-        && apt-get install -y --no-install-recommends libnvinfer-dev=5.0.2-1+cuda10.0 \
-        && rm -rf /var/lib/apt/lists/*
+        && apt-get install -y --no-install-recommends libnvinfer5=5.0.2-1+cuda${CUDA} \
+        && apt-get clean \
+        && rm -rf /var/lib/apt/lists/*)
 
 # Configure the build for our CUDA configuration.
 ENV CI_BUILD_PYTHON python
@@ -40,9 +46,10 @@ ENV LD_LIBRARY_PATH /usr/local/cuda/extras/CUPTI/lib64:$LD_LIBRARY_PATH
 ENV TF_NEED_CUDA 1
 ENV TF_NEED_TENSORRT 1
 ENV TF_CUDA_COMPUTE_CAPABILITIES=3.5,5.2,6.0,6.1,7.0
-ENV TF_CUDA_VERSION=10.0
-ENV TF_CUDNN_VERSION=7
-
-# Check out TensorFlow source code if --build_arg CHECKOUT_TENSORFLOW=1
+ENV TF_CUDA_VERSION=${CUDA}
+ENV TF_CUDNN_VERSION=${CUDNN%%.*}
+# CACHE_STOP is used to rerun future commands, otherwise cloning tensorflow will be cached and will not pull the most recent version
+ARG CACHE_STOP=1
+# Check out TensorFlow source code if --build_arg CHECKOUT_TF_SRC=1
 ARG CHECKOUT_TF_SRC=0
-RUN test "${CHECKOUT_TF_SRC}" -eq 1 && git clone https://github.com/tensorflow/tensorflow.git /tensorflow_src
+RUN test "${CHECKOUT_TF_SRC}" -eq 1 && git clone https://github.com/tensorflow/tensorflow.git /tensorflow_src || true
diff --git a/tensorflow/tools/dockerfiles/partials/ubuntu/nvidia.partial.Dockerfile b/tensorflow/tools/dockerfiles/partials/ubuntu/nvidia.partial.Dockerfile
index a6393a3280c..1d40ed5f981 100644
--- a/tensorflow/tools/dockerfiles/partials/ubuntu/nvidia.partial.Dockerfile
+++ b/tensorflow/tools/dockerfiles/partials/ubuntu/nvidia.partial.Dockerfile
@@ -1,29 +1,34 @@
-FROM nvidia/cuda:10.0-base-ubuntu${UBUNTU_VERSION} as base
+ARG ARCH=
+ARG CUDA=10.0
+FROM nvidia/cuda${ARCH:+-$ARCH}:${CUDA}-base-ubuntu${UBUNTU_VERSION} as base
+ARG CUDNN=7.4.1.5-1
 
+# Needed for string substitution 
+SHELL ["/bin/bash", "-c"]
 # Pick up some TF dependencies
 RUN apt-get update && apt-get install -y --no-install-recommends \
         build-essential \
-        cuda-command-line-tools-10-0 \
-        cuda-cublas-10-0 \
-        cuda-cufft-10-0 \
-        cuda-curand-10-0 \
-        cuda-cusolver-10-0 \
-        cuda-cusparse-10-0 \
-        libcudnn7=7.4.1.5-1+cuda10.0 \
+        cuda-command-line-tools-${CUDA/./-} \
+        cuda-cublas-${CUDA/./-} \
+        cuda-cufft-${CUDA/./-} \
+        cuda-curand-${CUDA/./-} \
+        cuda-cusolver-${CUDA/./-} \
+        cuda-cusparse-${CUDA/./-} \
+        curl \
+        libcudnn7=${CUDNN}+cuda${CUDA} \
         libfreetype6-dev \
         libhdf5-serial-dev \
-        libpng12-dev \
         libzmq3-dev \
         pkg-config \
         software-properties-common \
         unzip
 
-RUN apt-get update && \
-        apt-get install nvinfer-runtime-trt-repo-ubuntu1604-5.0.2-ga-cuda10.0 \
+RUN [ ${ARCH} = ppc64le ] || (apt-get update && \
+        apt-get install nvinfer-runtime-trt-repo-ubuntu1604-5.0.2-ga-cuda${CUDA} \
         && apt-get update \
-        && apt-get install -y --no-install-recommends libnvinfer5=5.0.2-1+cuda10.0 \
+        && apt-get install -y --no-install-recommends libnvinfer5=5.0.2-1+cuda${CUDA} \
         && apt-get clean \
-        && rm -rf /var/lib/apt/lists/*
+        && rm -rf /var/lib/apt/lists/*)
 
 # For CUDA profiling, TensorFlow requires CUPTI.
 ENV LD_LIBRARY_PATH /usr/local/cuda/extras/CUPTI/lib64:$LD_LIBRARY_PATH
diff --git a/tensorflow/tools/dockerfiles/spec.yml b/tensorflow/tools/dockerfiles/spec.yml
index 19d96e7a3df..d19b1d15fc8 100644
--- a/tensorflow/tools/dockerfiles/spec.yml
+++ b/tensorflow/tools/dockerfiles/spec.yml
@@ -56,6 +56,13 @@ releases:
             - "{ubuntu}{jupyter}"
             - "{ubuntu-devel}{jupyter}"
 
+    ppc64le-dockerfiles:
+        is_dockerfiles: true
+        upload_images: false
+        tag_specs:
+            - "{ubuntu-ppc64le}{jupyter}"
+            - "{ubuntu-devel-ppc64le}{jupyter}"
+
 slice_sets:
 
     py:
@@ -122,6 +129,70 @@ slice_sets:
           args:
               - CHECKOUT_TF_SRC=1
 
+    ubuntu-ppc64le:
+        - add_to_name: "-ppc64le"
+          dockerfile_exclusive_name: "cpu-ppc64le"
+          dockerfile_subdirectory: "ppc64le"
+          args:
+              - UBUNTU_VERSION=18.04
+          partials:
+              - ubuntu/version
+              - ubuntu/cpu
+              - ubuntu/python
+              - tensorflow-ppc64le
+              - shell
+        - add_to_name: "-gpu-ppc64le"
+          dockerfile_exclusive_name: "gpu-ppc64le"
+          dockerfile_subdirectory: "ppc64le"
+          args:
+              - UBUNTU_VERSION=18.04
+              - ARCH=ppc64le
+              - CUDA=10.0
+              - TF_PACKAGE=tensorflow-gpu
+          partials:
+              - ubuntu/version
+              - ubuntu/nvidia
+              - ubuntu/python
+              - tensorflow-ppc64le
+              - shell
+          tests:
+              - import-gpu.sh
+          test_runtime: nvidia
+
+    ubuntu-devel-ppc64le:
+        - add_to_name: "devel-ppc64le"
+          dockerfile_exclusive_name: "devel-cpu-ppc64le"
+          dockerfile_subdirectory: "ppc64le"
+          partials:
+              - ubuntu/version
+              - ubuntu/devel-cpu
+              - ubuntu/python
+              - ubuntu/bazelbuild
+              - shell
+          tests:
+              - build-cpu.sh
+          args:
+              - UBUNTU_VERSION=18.04
+              - CHECKOUT_TF_SRC=1
+        - add_to_name: "devel-gpu-ppc64le"
+          dockerfile_exclusive_name: "devel-gpu-ppc64le"
+          dockerfile_subdirectory: "ppc64le"
+          args:
+              - UBUNTU_VERSION=18.04
+              - ARCH=ppc64le
+              - CUDA=10.0
+              - LIB_DIR_PREFIX=powerpc64le
+              - CHECKOUT_TF_SRC=1
+          partials:
+              - ubuntu/version
+              - ubuntu/devel-nvidia
+              - ubuntu/python
+              - ubuntu/bazelbuild
+              - shell
+          tests:
+              - build-gpu.sh
+          test_runtime: nvidia
+
     nightly:
         - add_to_name: "nightly"
           partials:
diff --git a/tensorflow/tools/dockerfiles/tools.Dockerfile b/tensorflow/tools/dockerfiles/tools.Dockerfile
index e8929295a5e..a96b2578cba 100644
--- a/tensorflow/tools/dockerfiles/tools.Dockerfile
+++ b/tensorflow/tools/dockerfiles/tools.Dockerfile
@@ -17,7 +17,7 @@
 #
 # You can use this image to quickly develop changes to the Dockerfile assembler
 # or set of TF Docker partials. See README.md for usage instructions.
-FROM debian:stretch
+FROM ubuntu:16.04
 LABEL maintainer="Austin Anderson <angerson@google.com>"
 
 RUN apt-get update && apt-get install -y python3 python3-pip bash curl