Popular Deep Learning Frameworks

Deep learning frameworks provide data scientists, developers, and researchers a high-level programming language to architect, train, and validate deep neural networks.

TensorFlow*

Based on Python*, this deep learning framework is designed for flexible implementation and extensibility on modern deep neural networks. In collaboration with Google*, TensorFlow has been directly optimized for Intel® architecture to achieve high performance on Intel® Xeon® Scalable processors.

 

Downloads

Containers Binaries Source Code

 

TensorFlow is included as part of the Intel® AI Analytics Toolkit, powered by oneAPI.

PyTorch*

This Python package provides one of the fastest implementations of dynamic neural networks to achieve speed and flexibility. In collaboration with Facebook*, this popular framework is now combined with many Intel® optimizations to provide superior performance on Intel architecture, most notably Intel Xeon Scalable processors.

 

Downloads

Binaries Source Code

 

PyTorch is included as part of the Intel® AI Analytics Toolkit, powered by oneAPI.

 

Documentation

Get Started Guide

MXNet*

An open-source, deep learning framework, MXNet is highly portable, lightweight, and designed to offer efficiency and flexibility through imperative and symbolic programming. It includes built-in support for optimizations from Intel to achieve high performance on Intel Xeon Scalable processors.

 

Downloads

Containers Binaries Source Code

 

Documentation

Installation Guide

Optimization Techniques

Performance Tips for CPUs

Inference Performance on Intel Xeon Scalable Processors

PaddlePaddle*

This open-source deep learning Python* framework from Baidu is known for user-friendly, scalable operations. Built using Intel® Math Kernel Library for Deep Neural Networks, this popular framework provides fast performance on Intel Xeon Scalable processors as well as a large collection of tools to help AI developers.

 

Downloads

Source Code

 

Documentation

Get Started Guide

Caffe*

Created by the Berkeley Vision and Learning Center (BVLC) and community contributors, the Intel® Optimization for Caffe* is a fork maintained by Intel that is optimized for Intel architectures. This optimized branch of Caffe is one of the most popular frameworks for image recognition with improved performance on Intel Xeon Scalable processors.

 

Downloads

Containers Binaries Source Code

Intel® AI Analytics Toolkit

Accelerate end-to-end machine learning and data science pipelines with optimized deep learning frameworks and high-performing Python* libraries.

Get It Now

Explore the Intel® DevCloud

With built-in AI acceleration, this powerful cluster of the latest Intel Xeon Scalable processors provides the right environment for your experiments with code and debugging.

Webinar: Get Deep Learning Framework Performance

Learn how frameworks optimized by Intel can accelerate your AI applications on Intel architecture.