Daijiworld Media Network - Seattle

Seattle, Oct 20: Amazon.com is introducing a new software tool that allows developers to run code written in the popular artificial intelligence (AI) programming framework PyTorch on its custom Trainium AI chips without rewriting the code, in a move aimed at challenging Nvidia’s dominance in the AI computing market.

The tool, called TorchNeuron, enables developers to run PyTorch code on Amazon’s Trainium chips with a single configuration change, according to Amazon’s Annapurna Labs chip unit. The software is scheduled to be announced at a developer conference in Silicon Valley on Monday and will be released publicly as a free, open-source offering.

PyTorch, an open-source framework stewarded by Meta, is widely used by AI developers. Nvidia has maintained a strong position in the AI computing market through its CUDA software platform, which it has developed over more than a decade to enable AI applications to run efficiently on its chips.

Other cloud computing companies, including Google, are also developing tools to attract AI developers to their own custom-designed chips by improving support for PyTorch.

Colin Brace, vice-president of software at Annapurna Labs, said the initiative aims to simplify the process of adopting Amazon’s in-house AI chips by reducing the steps developers need to take to get started.

“This has been about meeting customers where they're at,” he told Reuters in an interview.

Brace said more than 30 early customers across sectors such as robotics, world models and voice AI were already using the software. He added that switching to Amazon’s chips could deliver substantial cost savings.