AI Chip Startup MatX Secures $500M to Challenge Nvidia
MatX, a semiconductor startup established by former Google hardware engineers, has secured $500 million in a Series B funding round. The investment aims to accelerate the development of specialized processors designed to outperform Nvidia’s dominant GPUs in the training and deployment of large language models (LLMs).
The financing round was co-led by Jane Street and Situational Awareness, an investment vehicle established by former OpenAI researcher Leopold Aschenbrenner. CEO and founder Reiner Pope confirmed the news on Tuesday via a LinkedIn update.
Strategic Backing and Market Ambitions
The startup attracted a diverse group of investors, signaling strong confidence in its hardware approach. Participants in the round include:
- Marvell Technology
- NFDG
- Spark Capital
- Patrick Collison and John Collison, co-founders of Stripe
MatX is positioning its technology to be ten times more efficient than current Nvidia offerings when handling LLM workloads. While the company has not disclosed its current valuation, the capital injection follows a Series A round of approximately $100 million led by Spark Capital in 2024, which previously pegged the firm’s valuation above $300 million.
Industry Context
The race to provide alternatives to Nvidia hardware is intensifying. MatX faces competition from players like Etched, which recently secured $500 million in funding at a $5 billion valuation.
The founders of MatX bring significant industry pedigree to the venture:
- Reiner Pope: Formerly led AI software development for Google’s proprietary TPU chips.
- Mike Gunter: Served as a lead designer for Google’s TPU hardware before co-founding the startup in 2023.
MatX plans to utilize the new capital to facilitate the manufacturing of its processors through TSMC. The company is currently targeting 2027 as the timeline for initial product shipments.