AMD has announced that it would be a founding member of the newly formed PyTorch Foundation. The organization, which will be part of the non-profit Linux Foundation, will encourage the adoption of Artificial Intelligence (AI) technology by promoting and supporting an ecosystem of open source projects using PyTorch, Meta’s Machine Learning (ML) software framework.
AMD, as a founding member, joins other industry leaders in prioritizing the future expansion of PyTorch’s robust community. AMD will assist the PyTorch Foundation by striving to democratize state-of-the-art tools, libraries, and other components to make these ML advances available to everyone, using inventions such as the AMD ROCm open software platform, AMD Instinct accelerators, Adaptive SoCs, and CPUs.
With its comprehensive product and software portfolio, AMD is well positioned to assist clients and partners in developing and deploying AI-powered applications ranging from the cloud and enterprise to the edge and endpoints. AMD can support a broad range of ubiquitous AI and ML models, from the tiny edge points to huge scale-out training and inference workloads, thanks to a diversified mix of hardware that includes AMD Instinct and Alveo accelerators, adaptable SoCs, and CPUs.
Additionally, AMD collaborates closely with the AI open community to promote and improve machine and deep learning capabilities and optimizations. Vitis AI offers a full AI inference development platform for AMD adaptive SoCs and Alveo data center accelerators. Vitis AI integrates with typical software development tools and makes use of a broad range of optimized open-source libraries to provide machine learning acceleration as part of software code.
To satisfy the demands of the AI/ML and HPC communities, the ROCm open software platform is continually expanding. Developers now have access to turn-key AI framework containers on AMD Infinity Hub, improved tools, simpler installation, and can anticipate lower kernel startup times and better application performance with the newest version of ROCm 5.0. In addition, AMD ROCm support has been upgraded from beta to stable with the new PyTorch 1.12 release.
