This Startup Just Raised $15 Million to Help AI Run on Any Chip, Not Just Nvidia's

Infinity, a young AI infrastructure company, has secured $15 million in funding at a $100 million valuation. The round was backed by Touring Capital, Principal VC, and individual researchers from OpenAI and Anthropic. The startup's mission: make it easier for AI models to run on chips other than Nvidia's.

Nvidia's dominance isn't just about hardware - it's also about CUDA, the software that lets developers write AI code that runs seamlessly on Nvidia GPUs. Most startups lack the resources to build custom low-level code, or "kernels," for alternative chips. Infinity aims to change that with a universal inference library that works across different chip types - SRAM, GPUs, phone chips, and Systolic Arrays. It's one of several startups chipping away at Nvidia's market grip, one product at a time.

Founded last year by former Google Brain researcher Jeremy Nixon, Infinity uses an AI agent called Ignition to automatically write, test, and optimize low-level code for non-Nvidia chips. The system learns and improves itself, adapting to proprietary architectures. Customers include chip maker D-Matrix, and the startup is in talks with other major chip and cloud companies. Rather than charging upfront license fees, Infinity takes a cut of the performance gains and cost savings it delivers. The company currently employs 26 people across design, operations, and engineering.

Source: TechCrunch