Super Micro Computer, Inc. (NASDAQ: SMCI) is now shipping NVIDIA Vera Rubin NVL72 racks that integrate its Data Center Building Block Solutions (DCBBS) and direct liquid cooling stack (DLC-2). CEO Charles Liang highlighted years of work on liquid‑cooling infrastructure, enabling customers to receive production‑ready, end‑to‑end systems. The NVIDIA Vera Rubin platform pairs AI GPUs with direct liquid cooling for unmatched throughput‑per‑watt, moving heat through cold plates, manifolds, and cooling towers.

Supermicro validates each rack with its full liquid cooling stack, accelerating deployment. A typical NVL72 rack holds 72 NVIDIA Rubin GPUs and 36 Vera CPUs across 18 compute trays, delivering 216 TB/s bandwidth and up to 54 TB of LPDDR5X memory. Cooling options include 1.8 MW in‑row CDUs with N+1 redundancy, rear‑door heat exchangers, and scalable DCBBS Blueprints that support power envelopes from 5 MW to gigawatt scale.

The solution offers balanced thermal, power, storage, and networking integration, backed by Supermicro’s end‑to‑end project management from site survey to ongoing support.