OLIX, a London-based AI chip startup founded by 25-year-old James Dacombe, has successfully raised over $250 million in its latest financing round. This funding is dedicated to the development and delivery of their proprietary Optical Tensor Processing Units (OTPUs). The company's goal is to build AI chips that are faster and more cost-effective than those currently offered by Nvidia, positioning itself as a significant challenger in the AI compute market.
OLIX believes its SRAM-architecture, integrated with photonics, can outperform HBM-based architectures in throughput per megawatt (throughput/MW) and total cost of ownership (TCO). Furthermore, they anticipate their technology will significantly surpass silicon-only SRAM-architectures in terms of interactivity and latency, which are critical factors for next-generation AI applications.
The company is scaling rapidly and actively recruiting across several global locations, including London, Austin, San Francisco, Toronto, and Bristol. They are seeking talent to work on what they describe as "the most important compute problem of the next decade." OLIX was founded in 2024 and currently employs between 90 and 100 people across six countries, including the United Kingdom, United States, Canada, Australia, and Romania.
James Dacombe, who started his first company at 18, is now building OLIX with the ambition to solve fundamental challenges in frontier AI deployment. The company's vision, outlined in its "OLIX Compute Manifesto," emphasizes its approach to winning in the "inference-era compute" landscape. This significant funding round, reported by the Financial Times, underscores investor confidence in OLIX's innovative approach to AI chip technology.
