At their Build digital developers conference on May 19, Microsoft announced a Public Preview for GPU in Windows Subsystem for Linux (WSL). WSL is a layer that enables executing Linux binaries on Microsoft Windows computing systems.
The announcement disclosed the collaboration with NVIDIA to deliver CUDA GPU-acceleration support to masses of Windows users. A live demonstration also showed seamless integration for installing NVIDIA software, and then running CUDA and advanced Linux-based AI frameworks like TensorFlow.
NVIDIA is offering AI expertise across CUDA, open source software libraries like RAPIDS, and pre-validated deep learning industry frameworks such as PyTorch and TensorFlow to enable users to implement their Linux workflows. This includes running containers, for GPU accelerated computing in Windows. Those users can range from AI experts, to novice developers, to students.
NVIDIA’s ‘CUDA on WSL’ elements of this preview deliver access to advanced Linux infrastructure which otherwise could not be run on Windows machines currently in enterprise, SMB, education, or personal use. Containerization simplifies the introduction of deep learning training and inference capabilities to these users. A wide variety of NVIDIA graphics cards and laptops empower new experimentation with GPU acceleration too.
Microsoft has also published a blog that includes more details about the announcement.