d-Matrix

d-Matrix develops the Corsair AI inference platform, a PCIe-based accelerator that places compute in memory, together with the JetStream interconnect and the Aviator software SDK. The company also describes a roadmap toward 3D memory-compute substrates.

The company sells inference hardware and software to enterprises and data centers deploying generative AI; its website lists sales roles and a Demo Cloud for testing. The website does not describe a detailed revenue model, and it states that the team has shipped over 100 million chips.

The company lists locations in Santa Clara (Silicon Valley), Sydney, Bengaluru, Toronto and Belgrade, and its careers page shows about 40 open roles, mostly in software, hardware and AI infrastructure engineering. Its investor materials cite a $275M Series C that brought total funding to $450M.

Product: Corsair AI inference accelerator card with JetStream interconnect and Aviator software SDK for generative AI inference.