The company’s optical technology is intended to help AI data centres to dramatically reduce costs and boost performance – while cutting energy consumption.
The round was led by VC investor Constructor Capital, and was also supported by existing investor IP Group, alongside new investors PhotonVentures, Journey Ventures, LIFTT, Qubits Ventures, State Farm Ventures, and TIS.
According to Lumai the funds raised will be used to support its next stage of growth – advancing product development, doubling its headcount, and expanding its presence in the US.
Power-hungry AI models and the drive to artificial general intelligence are pushing silicon-based compute to its limits and data centre power use in the US is expected to triple by 2028, consuming up to 12% of the country’s power.
Spun out of the University of Oxford, Lumai is looking to address the limitations of AI compute by using optical processing to accelerate large language models (LLMs) and other transformer-based AI. Its technology processes AI’s core arithmetic operations within optical beams traveling through 3D space, bypassing the limits of silicon GPUs and integrated photonics.
By utilising low-cost optical components, the PCIe form factor enables cost-effective, high-performance AI inference, helping to position Lumai as a key enabler in the fast growing AI market.
According to Lumai, its accelerator will cut AI inference costs to one-tenth of today’s top solutions and its unique design will deliver 50 times the performance of silicon-only accelerators while using just 10% of the power required for AI in data centres.
Tim Weil, CEO and co-founder at Lumai, commented, “The future of AI demands radical breakthroughs in computing. The cost of current LLMs is unsustainable, and next-generation AI won’t happen without a major shift. Lumai’s optical computing design overcomes the scalability challenges that have held others back and dramatically reduces power consumption, which will drive down the cost of AI.”
