d-Matrix Adopts NVIDIA NVLink Fusion for Rack-Scale XPU Deployment

d-Matrix will integrate its Raptor XPUs with NVIDIA NVLink Fusion, enabling rack-scale deployment on the NVIDIA MGX architecture. The blog details the high‑bandwidth, low‑latency connection and shared cooling and power infrastructure, offering customers a faster, lower‑risk path to scale ultralow‑latency inference. No performance numbers are provided.