Blog

TorchTitan is a PyTorch-native platform for rapid experimentation and large-scale pre- and post-training of generative…

PyTorch users on AMD MI350-series GPUs can now use FlyDSL through existing operator APIs for…

TL;DR TorchTPU is an open source collaboration between Google and Meta that introduces a native…

TL;DR Torch-spyre’s Inductor decides how each kernel’s work is divided across Spyre’s cores and manages…

Frontier models span language, video and audio, on architectures from dense and mixture-of-experts transformers to…

Why Zephon? The data that a model sees during training is one of the most…

TL;DR Deploying PyTorch models on microcontrollers has historically required converting to other framework formats, maintaining…

TL;DR The PyTorch-based software stack for IBM's Spyre accelerator recently moved its torch.compile backend from…

TL;DR Agentic workloads change the traffic an inference server sees. Unlike single-turn chat, an agent…

TL;DR Spyre becomes a native PyTorch device by connecting PyTorch's existing device, allocator, stream, and…

This post explores the FBTriton kernel design for Table Batched Embedding (TBE) forward and backward…

TL;DR If you need to decode or encode media, whether it's images, video, or audio,…

TL;DR The Accelerator Integration Working Group plays a vital role in standardizing how new hardware…