Nvidia challenger AI chip startup MatX raised $500M

MatX, an AI chip startup founded by former Google TPU engineers, has secured a $500 million Series B funding round to challenge Nvidia's control over the AI hardware market. The round, led by Jane Street and former OpenAI researcher Leopold Aschenbrenner's fund, Situational Awareness, signals continued investor confidence in specialized hardware designed to accelerate large language model training and inference.

The new capital will be used to manufacture its chips with TSMC, with a target to begin shipping in 2027, according to CEO Reiner Pope. The company's founders bring direct experience from Google's proprietary AI chip efforts; Pope led AI software development for Google's TPUs, while co-founder Mike Gunter was a lead hardware designer for the project. This round follows a Series A of approximately $100 million in 2024. Other investors in the Series B include Marvell Technology, Spark Capital, and Stripe co-founders Patrick and John Collison.

MatX aims to develop processors that are 10 times more efficient than Nvidia's current GPUs for both training and running large language models. The significant funding underscores the high-stakes competition in the AI semiconductor space, where multiple startups are racing to provide alternatives to Nvidia's established ecosystem. The company's closest competitor, Etched, recently raised a similar $500 million round at a reported $5 billion valuation, indicating a strong market appetite for new AI hardware architectures.

The parallel $500M funding rounds for MatX and its competitor Etched demonstrate that the market is not betting on a single 'Nvidia killer,' but is instead capitalizing multiple well-credentialed teams in a high-stakes race to build specialized AI silicon.