AWS releases TorchNeuron so PyTorch apps can run on Trainium
Original title: Amazon is taking its biggest shot at Nvidia yet with new TorchNeuron dev tool
Amazon says one configuration line moves PyTorch apps to Trainium and it will upstream changes to loosen Nvidia ties in PyTorch.
Amazon Web Services released TorchNeuron, a PyTorch-based tool that it says runs existing PyTorch apps on its Trainium chip after a single configuration change, with no code rewrite. AWS argues Trainium covers both training and inference inside its cloud and cites Deepgram processing 70% more audio per dollar after moving with native PyTorch. Amazon also says it will contribute upstream PyTorch changes meant to remove code that ties the framework to specific chips, including Nvidia’s default position via CUDA.
Unconfirmed: The 85% PyTorch figure, cost claims, and Deepgram result are Amazon’s statements.
- TorchNeuron is adapted from PyTorch and needs one configuration change, Amazon says.
- Amazon says about 85% of AI apps are written with PyTorch.
- AWS cites Deepgram processing 70% more audio per dollar on Trainium.
- Amazon plans to contribute hardware-agnostic changes to the main PyTorch project.
- PyTorch has been maintained by the Linux Foundation since 2022.
- Vendors
- Amazon Web ServicesNvidia
- Products
- TorchNeuronTrainiumPyTorchCUDA
- Organizations
- Linux FoundationAnnapurna Labs
From the original article
3 min read at techcrunch.com
Amazon wants to break Nvidia’s hold on PyTorch, the free toolkit used to build most AI apps, so that more developers run their apps on alternative chips like its own.
By Julie Bort · techcrunch.com
A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.