OLIX Computing Scores $312M for Photonic AI Breakthrough

UK-based startup OLIX Computing has secured $312 million in funding to scale its frontier inference technology, sending shockwaves through the crypto and AI communities. The Series B round values the company at a staggering $3.3 billion, up from just $1 billion mere months ago.

OLIX makes Optical Tensor Processing Units (OTPU), chips that use light instead of electricity to process AI math. This approach is faster, cheaper, and more energy-efficient than traditional GPUs from Nvidia, which rely on electrons flowing through silicon.

The company's innovative architecture eliminates the need for high-bandwidth memory (HBM), a costly and supply-constrained component that makes Nvidia's data center GPUs so expensive. Instead, OLIX envisions the data center as a 'token factory', purpose-built for churning out AI inference at scale.

Funded by investors including Hummingbird Ventures and Vertex Ventures, OLIX plans to expand into the US and Canada with a focus on 'AI infrastructure sovereignty'. The company's photonic approach is distinct from other startups like Cerebras and Groq, which pitch alternative chip architectures but still rely on similar memory architectures as Nvidia.

The $3.3 billion valuation after just 15 months of existence is aggressive by any standard, raising questions about the feasibility of photonic computing at scale. While academic labs have demonstrated the physics, manufacturing photonic chips with competitive yields remains an engineering challenge.