MatX, a startup developing chips tailored for large language models (LLMs), has closed an $80 million Series A funding round, according to multiple sources. The investment comes less than a year after the company secured $25 million in seed funding. Spark Capital led the round, placing MatX’s valuation at approximately $200 million pre-money and over $300 million post-money, according to a source who reviewed the deal.
MatX was co-founded two years ago by Mike Gunter and Reiner Pope, both alumni of Google’s Tensor Processing Unit (TPU) team. Gunter contributed to TPU hardware design, while Pope focused on AI software development. Drawing on their experience, the duo aims to address the global shortage of chips optimized for AI workloads.
According to MatX’s website, the company’s chips are specifically designed for AI tasks involving “at least” 7 billion parameters, with ideal performance at 20 billion or more. The chips are engineered for scalability in large clusters, leveraging an advanced interconnect system that facilitates efficient communication between AI processors.
Gunter and Pope outlined their vision in an interview with Bloomberg, stating their ambition to produce chips that are 10 times more efficient than Nvidia’s GPUs in training LLMs and delivering AI results. The startup’s advanced technology and affordability aim to position MatX as a competitor in the growing AI chip market.
MatX’s seed round in December 2023 was led by prominent AI angel investors Nat Friedman, former GitHub CEO, and Daniel Gross, who once led Apple’s AI and search initiatives after Apple acquired his startup, Cue. Gross has since co-founded an AI company, Safe Superintelligence, alongside former OpenAI chief scientist Ilya Sutskever.
The Information previously reported that MatX was seeking to raise between $75 million and $100 million for its Series A, indicating strong investor interest amid the AI boom. The startup joins a growing list of AI-focused chip companies attracting significant funding. For instance, Groq, founded by former TPU engineer Jonathan Ross, saw its valuation soar to $2.8 billion in August, nearly tripling since April.
The surge in interest in AI technologies has spurred unprecedented demand for processors capable of handling large-scale workloads. Companies like MatX aim to fill this gap by offering cutting-edge solutions that combine performance, scalability, and cost-efficiency. With its latest funding, MatX is poised to make significant strides in addressing the needs of the AI industry.
