Top AI Product

Qualcomm is buying Modular in an all-stock deal worth about $3.9 billion — and the target isn’t a chip, it’s software. Modular makes the Mojo programming language and the MAX inference platform, the closest thing the industry has to an open answer to Nvidia’s CUDA lock-in. That’s what Qualcomm actually paid for.

## What Modular built

MAX lets a trained AI model run across CPUs, GPUs, NPUs, and custom accelerators without rewriting the app for each target. Mojo is the language underneath it, built to give Python-style ergonomics with systems-level performance. Together they’re a portability layer: write once, deploy on whatever silicon is cheapest or fastest, instead of being married to one vendor’s stack.

## Why the deal matters

Nvidia’s real moat was never just the hardware — it was CUDA, the software everyone’s models are written against. Qualcomm bolting Modular onto its own silicon is a bet that the moat can be drained from the software side. For developers, the pitch is performance-per-watt and one toolchain spanning device, edge, and cloud. The deal is expected to close in the second half of 2026, pending regulators.