NNPACK

NNPACK is an acceleration package for neural network computations. NNPACK aims to provide high-performance implementations of convnet layers for multi-core CPUs.

NNPACK is not intended to be directly used by machine learning researchers; instead it provides low-level performance primitives to be leveraged by higher-level frameworks, such as Torch, Caffe, Tensorflow, Theano, and Mocha.jl.

Requirements

Host system

Linux or OS X host system
x86-64 processor with AVX2 instruction set
- NNPACK is optimized for Intel Skylake, but can run on Haswell & Broadwell processors too
- SSE2 instruction set can be targeted using --enable-psimd configuration option, but for performance reasons it is not recommended for production use

Cross-compilation options:

Native Client (x86-64) to run as a packaged Google Chrome App
Portable Native Client to run inside Google Chrome (no packaging required)
Android with x86/x86-64 (SSE2), ARMv7 with NEON, or ARM64 architecture

Features

Fast convolution algorithms based on Fourier transform and Winograd transform.

Forward propagation performance on Intel Core i7 6700K vs BVLC Caffe master branch as of March 24, 2016 (protobufs from convnet-benchmarks, integration via caffe-nnpack):

Library	Caffe	NNPACK	NNPACK	NNPACK
Algorithm	im2col + sgemm	FFT-8x8	FFT-16x16	Winograd F(6x6, 3x3)
AlexNet:conv2	315 ms	129 ms	86 ms	N/A
AlexNet:conv3	182 ms	87 ms	44 ms	70 ms
AlexNet:conv4	264 ms	109 ms	56 ms	89 ms
AlexNet:conv5	177 ms	77 ms	40 ms	64 ms
VGG-A:conv1	255 ms	303 ms	260 ms	404 ms
VGG-A:conv2	902 ms	369 ms	267 ms	372 ms
VGG-A:conv3.1	566 ms	308 ms	185 ms	279 ms
VGG-A:conv3.2	1091 ms	517 ms	309 ms	463 ms
VGG-A:conv4.1	432 ms	228 ms	149 ms	188 ms
VGG-A:conv4.2	842 ms	402 ms	264 ms	329 ms
VGG-A:conv5	292 ms	141 ms	83 ms	114 ms
OverFeat:conv2	424 ms	158 ms	73 ms	N/A
OverFeat:conv3	250 ms	69 ms	74 ms	54 ms
OverFeat:conv4	927 ms	256 ms	272 ms	173 ms
OverFeat:conv5	1832 ms	466 ms	524 ms	315 ms

Built-in expert-tuned kernels with very high performance:
- Fast Fourier transform
- Winograd transform
- Matrix-matrix multiplication (GEMM)
- Matrix-vector multiplication (GEMV)
- Max-pooling.
Multi-threaded SIMD-aware implementations of neural network layers.
Implemented in C99 and Python without external dependencies.
Extensive unit tests using C++ and Google Test.
Supports Native Client target and outperforms native Caffe/CPU when running inside Chrome.

Layers

Convolutional layer
- Training-optimized forward propagation (nnp_convolution_output)
- Training-optimized backward input gradient update (nnp_convolution_input_gradient)
- Training-optimized backward kernel gradient update (nnp_convolution_kernel_gradient)
- Inference-optimized forward propagation (nnp_convolution_inference)
Fully-connected layer
- Training-optimized forward propagation (nnp_fully_connected_output)
- Inference-optimized forward propagation (nnp_fully_connected_inference)
Max pooling layer
- Forward propagation, both for training and inference, (nnp_max_pooling_output)
ReLU layer (with parametrized negative slope)
- Forward propagation, both for training and inference, optionally in-place, (nnp_relu_output)
- Backward input gradient update (nnp_relu_input_gradient)
Softmax layer
- Forward propagation, both for training and inference, optionally in-place (nnp_softmax_output)

Building

NNPACK can be build on OS X and Linux.

Download, build and install PeachPy

git clone https://github.com/Maratyszcza/PeachPy.git
cd PeachPy
[sudo] pip install --upgrade -r requirements.txt
python setup.py generate
[sudo] pip install --upgrade .

Install ninja build system and ninja-syntax Python module

sudo apt-get install ninja-build || brew install ninja
[sudo] pip install ninja-syntax

Then clone and build NNPACK itself

git clone --recursive https://github.com/Maratyszcza/NNPACK.git
cd NNPACK
python ./configure.py
ninja

You can optionally add --enable-shared argument for configure.py to additionally build NNPACK as shared library (.so or .dylib). Shared library configuration is not recommended unless you need to load and use NNPACK through some FFI interface (e.g. Lua's ffi module or Python's ctypes module).

Cross-compilation for Native Client

Download and setup Native Client SDK
Set NACL_SDK_ROOT variable to a versioned SDK directory (e.g. ~/nacl_sdk/pepper_49).
Configure NNPACK with --host=x86_64-nacl-glibc or --host=x86_64-nacl-newlib (recommended) option.

Cross-compilation for Portable Native Client

Download and setup Native Client SDK
Set NACL_SDK_ROOT variable to a versioned SDK directory (e.g. ~/nacl_sdk/pepper_49).
Configure NNPACK with --host=pnacl-nacl-newlib option.

Cross-compilation for Android

Download and setup Android NDK
Add ndk-build to PATH variable
Navigate to NNPACK directory and build NNPACK with ndk-build build system. Note: an NDK project must be built with Clang toolchain to use NNPACK as a static libray.

Testing

NNPACK contains extensive test suite for transformation and neural network layers.

After configuration type ninja -t targets and choose the unit test that matches your subsystem of interest.

Packaging

Binary packages need to distribute two files: include/nnpack.h and lib/libnnpack.a (also lib/libnnpack.so or lib/libnnpack.dylib if NNPACK was configured with shared library support).

Bindings

Deep Learning Frameworks

Caffe2 natively supports NNPACK
nnpack-pr branch in ajtulloch/caffe - new integration of NNPACK (convolutional, fully-connected, and max-pooling layers) into Caffe.
Maratyszcza/caffe-nnpack - older and unmaintained integration of NNPACK (convolutional layers only) into Caffe.
szagoruyko/nnpack.torch - integration of NNPACK into Torch via ffi
tiny-dnn - header-only deep learning framework in C++11, which natively supports NNPACK.
MXNet - integration of NNPACK is being discussed in Issue #2986
See also discussion in Issue #1

Languages and Environments

node-nnpack - Node.js bindings
peterhj/libnnpack - Rust bindings

Users

Facebook uses NNPACK in production.
Prisma uses NNPACK in the mobile app.

Acknowledgements

The library is developed by Marat Dukhan of Georgia Tech with extensive advice from Nicolas Vasilache and Soumith Chintala of Facebook Artificial Intelligence Research. Andrew Tulloch of Facebook Artificial Intelligence Research contributed Caffe integration. We thank Andrew Lavin for fruitful discussions on Winograd transform-based implementations. NNPACK is a research project at Richard Vuduc's HPC Garage lab in the Georgia Institute of Technology, College of Computing, School of Computational Science and Engineering.

This material is based upon work supported by the U.S. National Science Foundation (NSF) Award Number 1339745. Any opinions, findings and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect those of NSF.

leizi007 / NNPACK