Olix has now raised over $530 million in total funding, following a $220 million Series A in February 2026 that valued the company at approximately $1 billion .
Alongside the funding, Olix made two high-profile hires:
Olix is not building another GPU. The company's platform is fundamentally different from Nvidia's approach, targeting the specific bottleneck it believes is holding back AI inference: memory bandwidth .
Olix's architecture is built around two key components:
Olix says its approach could fundamentally change how AI models are deployed, especially for inference workloads where power efficiency and memory bandwidth are critical.
Olix plans to deliver its first products to customers in 2027, with first customer access targeted for the second half of the year . The company will use Series B proceeds to:
Olix is also expanding its workforce across London, Bristol, Austin, Toronto, and San Francisco, hiring for teams working on silicon, photonics, compilers, and systems .
Olix's rapid valuation growth — from $1 billion to $3.3 billion in six months — reflects intense investor appetite for alternatives to Nvidia's near-monopoly in AI chips. The company's photonic approach represents a bet that the future of AI inference will be defined not by raw compute throughput, but by how efficiently data moves between processors .
However, it's worth noting that Olix's performance claims are company-reported and not yet validated in production hardware . The first customer deliveries in 2027 will be a critical test of whether photonic computing can actually compete with established GPU architectures in real-world deployments.