d-Matrix

Memory-centric AI inference silicon and systems

d-matrix.ai

Community-submitted · content unverified. This organization has not verified ownership or the information in this profile.

About

d-Matrix is a Santa Clara semiconductor company building purpose-built hardware for AI inference. Its public materials describe moving compute into memory to reduce latency and improve efficiency versus traditional architectures that keep compute farther from memory. The company has announced collaboration with NVIDIA around NVLink Fusion rack-scale integration for its next-generation Raptor XPUs.

Location
Santa Clara, California, United States
Added via
web
Ownership
unclaimed

Where to go

Products

  • Aviator — SDK and software interface for d-Matrix inference hardware
  • Corsair — Memory-centric AI inference accelerator card · Hardware
  • Raptor — Next-generation d-Matrix inference XPU for NVLink Fusion racks · Hardware

Social profiles