Amazon Web Services (AWS) announced on Monday the release of TorchNeuron, a new development tool. This tool is intended to simplify the process for AI startups to move their AI applications onto Amazon’s Trainium AI chip.
TorchNeuron is compatible with PyTorch, an open-source toolkit used by developers to build, train, and run AI models. According to Amazon, the tool enables AI apps written with PyTorch to operate on Trainium by altering just one line of configuration, without requiring a code rewrite.
Amazon states that around 85% of AI applications are developed using PyTorch. Colin Brace, VP of Technology Software at AWS Annapurna Labs, noted that Trainium offers "price-performance plus the fact that it’s one platform for both training and inference, which improves the economics across a full workload."
Beyond offering an easy migration option, Amazon is also working to make PyTorch less reliant on specific hardware. The company plans to contribute code to the main PyTorch project with the goal of making the framework more hardware-agnostic for the broader developer community.