How-toSelected

Chris Lattner Explains MAX and Mojo for PyTorch Chip Support

Why we picked it

The non-obvious point is that Qualcomm is backing an open shared-stack approach to AI accelerators, echoing how LLVM standardised compiler support across vendors. Engineers evaluating non-GPU accelerators and chip vendors building toolchains should watch the keynote to judge whether MAX and Mojo can replace in-house stack development. Compared with vendor-specific SDKs like CUDA, the pitch targets portability across PyTorch deployments.

At PyTorchCon North America 2026, Chris Lattner, now at Qualcomm, presented a keynote titled "MAX and Mojo: The Best Way to Extend PyTorch", outlining how a shared software stack can bring PyTorch support to new AI chips. In a clip shared by the official PyTorch account, he argues that hardware vendors should build on common compiler infrastructure rather than each writing bespoke software for their silicon.

Body

Lattner draws a parallel with the history of compilers: GCC and earlier LLVM work allowed many hardware vendors to share one infrastructure instead of duplicating effort. He positions Mojo as a systems programming language in the Python family, and MAX as a stack providing graph compilers, kernels, drivers, and memory management for accelerators — the pieces a chip vendor traditionally has to build in-house to run PyTorch workloads.

The implication is that emerging AI chipmakers could adopt MAX and Mojo to reach PyTorch compatibility faster, similar to how LLVM lowered the barrier for new CPU and GPU architectures. For developers, the pitch is portable PyTorch code across a wider range of accelerator hardware.

Original · PyTorch

How can a shared software stack help bring PyTorch support to new chips? Chris Lattner (@clattner_llvm) (@Qualcomm) explains the approach behind MAX and Mojo in this clip from #PyTorchCon North America 2026.

Chris traces how GCC and LLVM let hardware vendors build on shared compiler infrastructure, then makes the case for a similar approach to AI chips. He describes Mojo as a systems programming language in the Python family and MAX as providing graph compilers, kernels, drivers, and memory management for accelerators.

Watch the full keynote, “MAX and Mojo: The Best Way to Extend PyTorch”: https://t.co/41hV5zEIhX

Industry Coordinates & Source Verification

This report is categorized under AITOP's "How-to" desk, focusing on US frontier AI labs and Silicon Valley technology leaders. Within our cross-regional analytical framework, it directly maps to AI developer toolchains, coding agents, and engineering workflows. Key entities and technical topics under continuous tracking include PyTorch, Mojo, MAX, Chris Lattner.

Source attribution: curated from PyTorch (@PyTorch).