Supermicro SuperServer SYS-F629P3-RC1B (Liquid-Cooled)

The Supermicro SuperServer SYS-F629P3-RC1B is a 2U dual-socket rack server engineered around factory-integrated Direct Liquid Cooling (DLC) technology, purpose-built to support the next generation of high thermal design power processors and accelerators without relying on traditional air-cooling infrastructure alone. The DLC system routes temperature-controlled coolant directly to cold plates mounted on the CPUs, memory, and other high-heat components, dramatically reducing the thermal resistance between silicon and the cooling loop. This architecture allows the system to sustain full-performance workloads in facilities operating at elevated ambient temperatures or in dense rack deployments where air cooling headroom is limited.

Designed for the Intel Xeon Scalable processor family in a dual-socket configuration, the SYS-F629P3-RC1B delivers substantial compute density per rack unit, supporting up to 400W TDP per CPU socket when paired with the liquid cooling subsystem. The platform accommodates DDR4 registered ECC memory across up to 24 DIMM slots, providing ample bandwidth and capacity for memory-intensive analytics, virtualization, and AI inference workloads. Multiple PCIe expansion slots support high-bandwidth GPU accelerators, NVMe co-processors, or high-speed networking cards, making this chassis a flexible foundation across a wide range of HPC and enterprise acceleration scenarios.

What distinguishes the SYS-F629P3-RC1B within Supermicro's FatTwin and liquid-cooled portfolio is its datacenter operational efficiency profile. By transferring heat to a facility water loop rather than exhausting it as hot air into the datacenter aisle, the server enables organizations to reduce Power Usage Effectiveness (PUE) ratios, lower CRAC unit load, and qualify for higher-density rack configurations that would otherwise be thermally constrained. The platform ships with Supermicro's IPMI 2.0-compliant Baseboard Management Controller and is compatible with Supermicro's SuperDoctor 5 management software, providing granular monitoring of coolant flow, temperatures, and system health through standard out-of-band interfaces.

Ideal for

• High-performance computing (HPC) clusters requiring sustained full-socket TDP operation across dense multi-rack deployments without proportional increases in CRAC cooling capacity
• AI and machine learning training infrastructure where high-TDP GPUs are co-located with processor sockets and collective thermal output exceeds conventional air-cooling thresholds
• Financial services risk analytics and real-time quantitative modeling demanding consistent clock frequency under continuous 100% CPU utilization across extended computational runs
• Virtualization consolidation projects targeting aggressive VM-to-rack-unit ratios in datacenters with warm-aisle containment or limited raised-floor airflow availability
• Scientific research and simulation environments such as computational fluid dynamics, molecular dynamics, or seismic processing that maintain near-peak CPU utilization for days or weeks continuously
• Datacenter modernization initiatives converting legacy air-cooled rows to hybrid or fully liquid-cooled rows to achieve measurable PUE improvements and sustainability targets

Technical specifications

Available from Omnixon Global. Submit an RFQ and our team will confirm configuration and availability for your order.