Amazon Web Services (AWS) released a new developer tool called TorchNeuron on Monday, October 19, 2026, according to TechCrunch. The tool lets AI startups move their applications onto Amazon's Trainium AI chips by changing just one line of configuration, with no code rewrite required. It is Amazon's most direct challenge yet to Nvidia's long-standing dominance in AI hardware.
The announcement matters because most AI software is built with PyTorch, a free open source toolkit. Amazon says about 85% of AI apps are written with PyTorch, and Nvidia has spent years making sure its CUDA software supports those apps seamlessly. That has made Nvidia's hardware the default choice for AI developers, and TorchNeuron is designed to loosen that grip.
How TorchNeuron works with PyTorch
TorchNeuron is adapted from PyTorch itself, so developers who already know PyTorch do not need to learn a new system. According to Amazon, the tool works by letting PyTorch-based AI apps run on Trainium with a single configuration change. Amazon argues that porting to Trainium is not only simple but also gives startups access to the entire AWS cloud.
AWS is not the first company to offer an easy path to its chips. PyTorch users can already port their apps to AMD GPUs or even Apple's chips with a simple configuration change. Amazon's pitch goes further, claiming that Trainium handles both training and inference on the same chip, which it says improves the economics across a full workload.
Amazon's push to decouple PyTorch from Nvidia
AWS is also pushing to make PyTorch less centered on Nvidia. The company says it is contributing changes upstream to the main PyTorch project, working with the community to remove code that ties the framework to specific chips. The goal is to make PyTorch more hardware-agnostic for all developers, not just Trainium users.
Colin Brace, VP of Technology Software at AWS Annapurna Labs, told TechCrunch via email that the real story is removing the lock-in that kept developers from having a real choice at all. He also cited Deepgram, an AI voice app provider, which moved to Trainium using native PyTorch support and now processes 70% more audio per dollar at production-grade accuracy.
