Google targets Nvidia's dominance in AI chips by launching "TorchTPU," an initiative to enhance its Tensor Processing Units’ compatibility with PyTorch, the leading AI software framework. The move aims to ease adoption for developers. Google plans to collaborate with Meta, aiming to make TPUs more appealing to customers relying on PyTorch, while also considering open-sourcing parts of the software. This shift could significantly lower barriers for companies seeking alternatives to Nvidia’s GPUs.