diff --git a/.dockerignore b/.dockerignore
new file mode 100644
index 000000000..7763a51dc
--- /dev/null
+++ b/.dockerignore
@@ -0,0 +1 @@
+sift1M
\ No newline at end of file
diff --git a/Dockerfile b/Dockerfile
new file mode 100644
index 000000000..88eda521f
--- /dev/null
+++ b/Dockerfile
@@ -0,0 +1,36 @@
+FROM nvidia/cuda:8.0-devel-ubuntu16.04
+MAINTAINER Pierre Letessier <pletessier@ina.fr>
+
+RUN apt-get update -y
+RUN apt-get install -y libopenblas-dev python-numpy python-dev swig git python-pip wget
+
+RUN pip install matplotlib
+
+RUN cd /opt/ && git clone https://github.com/facebookresearch/faiss.git
+
+WORKDIR /opt/faiss
+
+RUN mv example_makefiles/makefile.inc.Linux ./makefile.inc
+
+RUN sed -i -e 's%^BLASLDFLAGS=/usr/lib64/libopenblas.so.0%# BLASLDFLAGS=/usr/lib64/libopenblas.so.0%g' -e 's%^# BLASLDFLAGS=/usr/lib/libopenblas.so.0$%BLASLDFLAGS=/usr/lib/libopenblas.so.0%g' makefile.inc
+
+RUN make tests/test_blas -j $(nproc) && \
+    make -j $(nproc) && \
+    make tests/demo_sift1M -j $(nproc)
+
+RUN make py
+
+RUN cd gpu && \
+    make -j $(nproc) && \
+    make test/demo_ivfpq_indexing_gpu && \
+    make py
+
+# RUN ./tests/test_blas && \
+#     tests/demo_ivfpq_indexing
+
+
+# RUN wget ftp://ftp.irisa.fr/local/texmex/corpus/sift.tar.gz && \
+#     tar xf sift.tar.gz && \
+#     mv sift sift1M
+
+# RUN tests/demo_sift1M
diff --git a/INSTALL b/INSTALL
index 160b0dd88..34ea56144 100644
--- a/INSTALL
+++ b/INSTALL
@@ -15,6 +15,8 @@ involved:
 
 Steps 2 and 3 depend on 1, but they are otherwise independent.
 
+Alternatively, all 3 steps above can be run by building a Docker image (see section "Docker instructions" below).
+
 General compilation instructions
 ================================
 
@@ -247,6 +249,31 @@ The auto-tuning example above also runs on the GPU. Edit
 
 to enable and run it.
 
+
+Docker instructions
+===================
+
+For using GPU capabilities of Faiss, you'll need to run "nvidia-docker"
+rather than "docker". Make sure that docker
+(https://docs.docker.com/engine/installation/) and nvidia-docker
+(https://github.com/NVIDIA/nvidia-docker) are installed on your system
+
+To build the "faiss" image, run
+
+  nvidia-docker build -t faiss .
+
+If you want to run the tests during the docker build, uncomment the
+last 3 "RUN" steps in the Dockerfile. But you might want to run the
+tests by yourself, so just run
+
+  nvidia-docker run -ti --name faiss faiss bash
+
+and run what you want. If you need a dataset (like sift1M), download it
+inside the created container, or better, mount a directory from the host
+
+  nvidia-docker run -ti --name faiss -v /my/host/data/folder/ann_dataset/sift/:/opt/faiss/sift1M faiss bash
+
+
 Hot to use Faiss in your own projects
 =====================================
 
diff --git a/README.md b/README.md
index 57e534325..c7e0a3fd3 100644
--- a/README.md
+++ b/README.md
@@ -12,7 +12,7 @@ The GPU implementation can accept input from either CPU or GPU memory. On a serv
 
 ## Building 
 
-The library is mostly implemented in C++, with optional GPU support provided via CUDA, and an optional Python interface. The CPU version requires a BLAS library. It compiles with a Makefile. See [INSTALL](INSTALL) for details.
+The library is mostly implemented in C++, with optional GPU support provided via CUDA, and an optional Python interface. The CPU version requires a BLAS library. It compiles with a Makefile and can be packaged in a docker image. See [INSTALL](INSTALL) for details.
 
 ## How Faiss works