Wayve, the London autonomous-driving startup behind the chip-agnostic AI Driver platform, closed a $60m Series D extension on 15 April 2026 from AMD, Arm and Qualcomm, pushing post-money valuation to $8.6bn and total funding to roughly $1.5bn.1 CNBC broke the round; the three chip architectures invested jointly rather than at different price points.
Wayve builds end-to-end neural networks that learn driving behaviour from sensor data rather than rule-based stacks. The company licenses its software to car-makers rather than building its own vehicles, and its stated pitch is hardware-agnosticism: the same model trains and runs on any of the major inference architectures. The round syndicate lends weight to that claim, since none of AMD, Arm or Qualcomm would fund a customer locked to a rival chip. For the three vendors, co-funding a European automotive customer hedges against Nvidia concentration and validates their automotive IP roadmaps simultaneously.
The timing sits alongside the Sovereign AI Unit cohort announcement and Nscale's $2bn build-out, both of which sit on the same three-architecture spine. That is not coincidence. Britain's AI hardware stack is consolidating around a specific alliance of non-Nvidia silicon, and Wayve is now the automotive reference customer validating that stack; Nscale is the infrastructure reference customer. For automotive OEMs choosing a driving-stack vendor, Wayve's funders shorten the qualification timeline: whatever chip they pick, Wayve's model already runs on it.
For founders building in adjacent verticals (robotics, industrial AI, edge inference), the takeaway is market structure rather than valuation. The commercial centre of gravity in UK AI infrastructure has tilted toward AMD, Arm and Qualcomm customers, and state policy has now followed. Founders on Nvidia-only stacks should expect harder conversations at DSIT and a slower path to SAIU-adjacent contracts.
