Supermicro expanded its Open Compute Project rack portfolio on October 13, 2026, adding tighter integration across power, cooling, networking and system management for high-density AI infrastructure. The company says its ORv3 and ORW racks use centralized 48V DC busbars instead of separate server power supplies and cabling, while its liquid-cooling stack is designed to support denser CPU and GPU deployments.
The cooling side is substantial. Supermicro says its rear-door heat exchangers span 10 kW to 120 kW, while a compact 4U in-rack CDU can provide up to 250 kW of cooling capacity. Those systems sit alongside direct-to-chip cooling intended to remove heat from processors before residual exhaust reaches the data hall.
Networking is part of the same rack design. A 2U co-packaged-optics switch is specified for 102.4 Tb/s across 128 × 800G ports, with 40–60 VDC busbar power and liquid cooling. The announcement extends the same rack-scale trend seen in NVIDIA’s open Vera Rubin rack designs, Supermicro’s Vera Rubin NVL72 shipments, and Cisco’s 51.2 Tb/s G302 networking push.
Supermicro explains its rack-scale and liquid-cooled AI infrastructure approach.
Supermicro previously showed the same DCBBS strategy publicly, including rack-scale systems and liquid-cooling infrastructure.
Supermicro’s official announcement says the goal is faster deployment across complete AI clusters. Independent market coverage from Newsquawk noted that portfolio announcements are useful confirmation of the infrastructure trend, while actual commercial impact still depends on order flow, shipment cadence and margins. The hardware specifications are vendor claims until independently validated in production deployments.
BitcoinVersus.Tech Editor’s Note:
We volunteer daily to ensure the credibility of the information on this platform is Verifiably True. If you would like to support to help further secure the integrity of our research initiatives, please donate using the QR code provided
For best QR-code scanning results, please scan this donation code from a desktop or larger screen. Mobile screens may prevent camera recognition.
Leave a Reply