Common Questions

Why did AMD invest in Wayve in 2026?

AMD co-invested $60m in Wayve's April 2026 Series D extension to embed AMD GPU architecture in Wayve's chip-agnostic AI Driver system, giving AMD a stake in the autonomous-driving inference market alongside Arm and Qualcomm.Source: Lowdown

How does AMD compete with Nvidia in AI chips?

AMD's Instinct MI300X GPU series competes directly with Nvidia's H100/H200 in AI training and inference. Cloud providers have adopted Instinct GPUs to reduce Nvidia dependency, and AMD has grown from near-zero AI market share in 2022 to a meaningful position by 2026.

What is AMD's Instinct MI300X?

The Instinct MI300X is AMD's flagship AI accelerator GPU, designed for large-scale training and inference workloads. It combines compute and high-bandwidth memory on a single package and competes directly with Nvidia's H100 for enterprise AI infrastructure.

Background

AMD joined Arm and Qualcomm in a $60m Series D extension into Wayve on 15 April 2026, securing AMD GPU architecture a position in the autonomous-driving inference stack at a moment when Nvidia dominates that space . Wayve's AI Driver is chip-agnostic, and AMD's inclusion alongside the two other architectures powering Nscale's £2bn UK compute build-out makes this an infrastructure-level alignment, not a passive venture bet.

Advanced Micro Devices, founded in 1969 and headquartered in Santa Clara, California, designs CPUs and GPUs for data-centre, consumer, and embedded markets. Its Instinct MI300X GPU series is the most credible alternative to Nvidia's H100/H200 in AI training and inference workloads. AMD's share of the AI accelerator market has grown from near zero in 2022 to a meaningful single-digit percentage by 2026, driven by large cloud providers seeking to reduce Nvidia dependency. The company is listed on Nasdaq and generated approximately $25bn in revenue in 2024.

The Wayve investment is AMD's most public autonomous-driving commitment. It positions AMD's Instinct architecture as a peer to Arm's Neoverse and Qualcomm's Snapdragon Ride in the UK sovereign AI hardware stack, ahead of the scale-out phase of the Nscale datacentre build where inference workloads will run in volume.