TNW reports that Google is discussing two AI chips with Marvell: a memory processing unit and an inference-focused TPU. No contract is signed yet, but the talks show how serving models, not just training them, is driving custom silicon strategy.
NVIDIA's Newsroom account said on X on March 31, 2026 that Marvell is joining NVLink Fusion to expand the NVIDIA AI ecosystem. The linked press release says the partnership combines Marvell custom XPUs, NVLink Fusion-compatible networking, silicon photonics collaboration, and a $2 billion NVIDIA investment in Marvell to support semi-custom AI infrastructure.