Amazon's new software can challenge NVIDIA's AI chip dominance
What's the story
Amazon is introducing a new tool, TorchNeuron, to run code written in PyTorch, the most widely used AI coding language, on its Trainium AI chips. The move is aimed at challenging NVIDIA's dominance in the AI computing market. The software requires only a single configuration change and no rewriting of code, according to Amazon's Annapurna Labs chip unit.
Impact
TorchNeuron is free and open-source
TorchNeuron is set to be launched as a free and open-source software at an upcoming developer conference in Silicon Valley.
The move comes as part of Amazon's strategy to make it easier for AI developers to try out its in-house chips.
"This has been about meeting customers where they're at," Colin Brace, VP of Software for Annapurna Labs, told Reuters.
Adoption
Over 30 customers already using new software
Over 30 early customers, including those in robotics, world models, and voice AI, are already using the new software.
Brace also noted that switching to Amazon's chips can lead to significant cost savings for these businesses.
The development of TorchNeuron is a major step by Amazon to take on NVIDIA's long-standing dominance in the AI chip market with its Cuda software tool for PyTorch code optimization.