Search
On the 10th local time, AI inference chip manufacturer d-Matrix announced its entry into the NVIDIA NVLink Fusion ecosystem, with its rack-level XPU system adopting NVIDIA's MGX universal rack reference architecture. Starting with 'Raptor,' d-Matrix's next-generation inference XPUs will be integrated with multiple NVIDIA products, enabling vertical and horizontal scalability. Additionally, NVLink Fusion partners will provide customized connectivity solutions for d-Matrix's XPUs, with racks based on the MGX platform featuring modular cable-free trays. By joining this ecosystem, d-Matrix can achieve greater scalability for rack-level solutions without complex design selections, allowing it to focus more on the development of AI XPUs themselves.
-
C114 Communication Network -
Communication Home
