The NVIDIA Deep Learning SDK accelerates widely-used deep learning frameworks such as Caffe, CNTK, TensorFlow, Theano and Torch as well as many other deep learning applications. Choose a deep learning framework from the list below and follow the instructions on the framework page to install and get started.
For access to NVIDIA optimized deep learning framework containers, visit the NVIDIA GPU CLOUD. Developers can get started with accelerated releases of popular frameworks including NVCaffe, Caffe2, Microsoft Cognitive Toolkit (CNTK), Digits, MXNet, PyTorch, TensorFlow, Theano, and Torch, for use on-premises or on AWS P3 instances. Sign up for an NGC account to get started.
Caffe supports cuDNN v5 for GPU acceleration.
Supported interfaces: C, C++, Python, MATLAB, Command line interfaceLearning Resources
Caffe2 is a deep learning framework enabling simple and flexible deep learning. Built on the original Caffe, Caffe2 is designed with expression, speed, and modularity in mind, allowing for a more flexible way to organize computation.
Caffe2 supports cuDNN v5.1 for GPU acceleration.
Supported interfaces: C++, PythonLearning Resources
Microsoft Cognitive Toolkit supports cuDNN v5.1 for GPU acceleration.
Supported interfaces: Python, C++, C# and Command line interfaceDownload CNTK
Theano supports cuDNN v5 for GPU acceleration.
Supported interfaces: Python
Torch supports cuDNN v5 for GPU acceleration.
Supported interfaces: C, C++, Lua
MXnet supports cuDNN v5 for GPU acceleration.
Supported Interfaces: Python, R, C++, JuliaDownload MXnet
Chainer supports cuDNN v5.1 for GPU acceleration.
Supported Interfaces: PythonDownload Chainer
cuDNN version depends on the version of TensorFlow and Theano installed with Keras.
Supported Interfaces: PythonDownload Keras
If you’re a framework developer and would like to see your framework listed here, please get in touch with us at firstname.lastname@example.org