The company aims to dismantle the exclusivity of Nvidia’s CUDA software, which currently traps developers within a specific hardware ecosystem. While Nvidia’s rise was fueled by the ability to turn graphics processors into general-purpose AI engines, Infinity is building a universal kernel-level software layer. This platform is designed to operate seamlessly across diverse hardware, from mobile chips to systolic arrays, effectively automating the creation of low-level code that usually requires months of manual engineering.
In section Startups & Technology
Infinity raises $15M to challenge Nvidia's software dominance
In a direct challenge to Nvidia’s grip on the AI hardware market, infrastructure startup Infinity has secured $15 million in funding at a $100 million valuation. The round attracted backing from Touring Capital and Principal VC, alongside individual researchers from industry giants OpenAI and Anthropic.
Founder Jeremy Nixon, a former Google Brain researcher and creator of the AGI House, developed an AI agent named Ignition to handle this heavy lifting. The system autonomously writes, tests, and optimizes code for non-Nvidia chips, self-correcting to ensure performance gains. By shifting from traditional licensing to a revenue model based on measurable performance improvements, Infinity is already working with emerging chipmakers like D-Matrix. The startup currently employs 26 people, betting that automated invention can replace the bottleneck of human-written hardware code.
Comments (0)
No comments yet. Be the first!