QuantaGrid D76V-1U
Enhanced Compute & HBM4 Bandwidth: Optimized compute and 288GB HBM4
memory per GPU to increase AI inferencing and training, integrated with almost 3x HBM4 bandwidth and 2x NVIDIA NVLink™ 6 bandwidth to elevate GPU-to-GPU communication efficiency with low latency.
Network Upgrades: New NVIDIA ConnectX®-9 SuperNIC™ cards deliver 1.6 Tb/s connectivity per GPU in NVIDIA Vera Rubin systems, with enhanced networking platforms for peak AI workload efficiency.
Optimized Serviceability: The NVIDIA Vera Rubin system with cableless design provides board to board connection through PHD2 connectors to deliver high speed signals and reduce cable routing effort. Along with modularized demonstration in chassis, NVIDIA Vera Rubin simplifies assembly and serviceability complexity and
exerts outstanding mechanical technique to the tailor-made system.
EfficientCooling:Liquid-cooled NVIDIA Vera Rubin NVL72 racks maximize
compute density and floor space utilization to align with carbon-free footprint and ESG goals, fulfilling 100% liquid cooling for the Vera Rubin NVL72 compute and switch trays.