Summary
The accelerator will combine LPDDR6, PCIe Gen5 and large-die technologies, with tape-out targeted for the first half of 2027. Global mass production is planned for 2028, with hyperscalers and cloud providers as target customers.
Why it matters
The contract is evidence that AI infrastructure demand is creating a growing market for custom inference silicon designed around specific workloads and power constraints.
Related coverage
- Anthropic IPO prospectus reveals unprecedented compute commitments and losses
- Samsung commits $1 billion to KKR-backed Helix AI infrastructure
- SiMa.ai raises $150 million to scale purpose-built physical-AI chips