d-Matrix to integrate Nvidia NVLink Fusion in AI inference chips

AI chip startup d-Matrix has announced it will integrate Nvidia’s NVLink Fusion technology into its upcoming Raptor processors. This integration allows d-Matrix hardware to plug directly into Nvidia data-center systems, aiming to capture market share in the growing AI inference sector.

While Nvidia’s high-end chips are widely used for model training, d-Matrix is focusing on inference—the daily deployment and execution of AI models. The Raptor chips are designed for high-speed, low-latency applications such as chatbots, voice agents, and coding assistants. To ensure rapid data flow, the Santa Clara-based company is also partnering with connectivity provider Astera Labs.

The Raptor processor is expected to complete its final design phase by the end of 2024, with compatible server racks scheduled for market release in 2027. d-Matrix, which has received backing from Microsoft, reached a $2 billion valuation following a $450 million capital raise last year.