PyTorch supports new AI accelerators
The story so far
PyTorch supports IBM's dataflow AI accelerator Spyre through torch-spyre and the Spyre vLLM plugin, becoming its first-class backend and supporting multi-accelerator scheduling. This enables models and workflows to run on the inference-optimized Spyre, suitable for IBM Z, LinuxONE, and Power systems. IBM claims that through the familiar PyTorch workflow, users can run models in a way similar to using Nvidia GPUs, aiming to make hardware selection an actual option rather than a software integration burden.
AI-written from the reports · Updated 23 days ago
Coverage timeline
Follow the coverage to see every side of the story.
- IBM Research BlogPyTorch opens the AI ecosystem to a new accelerator class
IBM and the PyTorch Foundation announce that Spyre accelerator will become the first-class backend for PyTorch and support multi-accelerator scheduling. The torch-spyre compiler, Spyre vLLM plugin, and continuous integration efforts enable models and workflows to run on Spyre, an architecture optimized for inference, suitable for IBM Z, LinuxONE, and Power systems.