Amazon Launches TorchNeuron to Break Nvidia's Grip on AI Chips

Amazon is taking the fight directly to Nvidia. The tech giant's chip unit, Annapurna Labs, has launched TorchNeuron — a new tool that lets developers run code written in PyTorch, the most popular AI programming framework, on Amazon's own Trainium AI chips with a single configuration change and zero rewriting.
The move strikes at the heart of Nvidia's moat. For more than a decade, Nvidia's CUDA software has made PyTorch code run best on its chips, locking most AI developers into its ecosystem. TorchNeuron, announced at a Silicon Valley developer conference and released as free open-source software, aims to collapse that barrier. Colin Brace, vice president of software for Annapurna Labs, told Reuters: 'This has been about meeting customers where they are at.' Amazon says more than 30 early customers — spanning robotics, world models and voice AI — are already using it, with steep cost savings reported.
Amazon is not alone: Google is also building better PyTorch tooling for its own custom AI chips as cloud giants race to give developers cheaper alternatives to Nvidia. With AI training and inference costs dominating tech budgets, the winner of the software battle may decide who controls the next generation of AI computing.
Source: read more
Comments
Post a Comment