D-Matrix

d-Matrix develops an AI inference computing platform for datacenters that speeds up generative AI workloads while cutting power use and infrastructure costs. Its Corsair inference accelerators and JetStream interconnect move data quickly between chips to keep response times low, even at high request volumes. SquadRack extends this stack to rack scale so operators can deploy and grow multimodal and agentic AI services without redesigning their data center.