Data Center Building Block Solutions®
Supermicro's Data Center Building Block Solutions (DCBBS) delivers complete, modular AI infrastructure. Built from validated components and sub-systems, DCBBS provides end-to-end deployment flexibility – from individual GPUs and networking switches to complete racks, site infrastructure, management software, and professional services.
Rapid Time-to-Deployment with First-to-Market
Total Solutions with Onsite Services
Customized to Your Workloads
Validated and Tested Before Shipping
Data Center Building Block Solutions® (DCBBS)
AI and Compute Systems
Industry's broadest AI and compute system portfolio with unmatched customization – from GPU systems to petascale storage, multi-node systems, and liquid cooling.
In-Rack Solutions
Complete in-rack integration: cooling distribution units & manifolds, hose kits, networking switches, power shelf, and battery backup unit, RDHx for maximum density and efficiency.
In-Row Solutions
Scalable liquid cooling from 200kW Liquid-to-Air Sidecars to 1.8MW In-Row CDUs, plus pre-validated SuperCluster solutions with integrated networking for turnkey multi-rack deployment.
Site Infrastructure Solutions
Site-level cooling infrastructure with scalable water and dry cooling towers for thermal and power management. Optimized cabling design and implementation for high-performance data pipeline.
Management Software Suite
End-to-end SuperCloud software from bare metal to AI workloads – delivering unified infrastructure control, deployment automation, developer tools, and multi-tenant GPU cloud management.
Services & Onsite Deployment
Comprehensive professional services from data center design and solution validation through on-site deployment and ongoing support, with 4-hour response options for mission-critical uptime.
AI and Compute Systems
Choose from the industry's broadest portfolio of AI and compute systems, powered by the latest technology from NVIDIA, AMD, and Intel. Supermicro’s Building Block approach delivers unmatched customization and optimization – from GPU-dense AI systems to petascale storage systems – allowing you to fine-tune every component for your workload. With liquid cooling cold plates that remove up to 98% of heat from critical electronics and seamless integration with leading software-defined networking partners, every system is optimized for maximum performance and efficiency from the ground up.
GPU and CPU Systems
Broad range of density, efficiency, and scalability-optimized systems featuring the latest AI and compute technology from NVIDIA, AMD, and Intel in various form factors designed for diverse workloads – from AI training and inference to high-performance computing.
Storage Systems
Petascale and object storage servers with seamless integration across leading software-defined networking partners, delivering maximum throughput and scalability for data-intensive AI workloads and large-scale AI model training and high-volume inference.
System Building Blocks, Cold Plates
Supermicro-designed liquid cooling cold plates that efficiently remove up to 98% of heat from critical electronics in systems, plus modular building block components that enable precise system customization and optimization for specific workload requirements.
4U NVIDIA HGX 8-GPU System
- Hot-swappable High-performance Drives
- CPU, DIMM, and PCIe Switch Cold Plates
- NVIDIA HGX B200 8-GPU
- GPU Cold Plates
- Power Supplies and High-speed NICs
In-Rack Solutions
Power, cool, and connect the most demanding AI and compute systems with purpose-built in-rack integration. Supermicro delivers Coolant Distribution Units supporting up to 250kW of liquid cooling capacity, high-performance networking switches including NVIDIA InfiniBand and Ethernet solutions. Combined with 33kW Power Shelves and Battery Backup Units that provide 90 seconds of 48V DC power for application checkpointing, Supermicro's In-Rack Solutions deliver the thermal management, networking fabric, power delivery, and resilience required for next-generation AI data centers – all optimized for maximum rack density and operational efficiency.
CDUs and CDMs
Coolant Distribution Units and Manifolds in various configurations, including vertical CDMs for higher server densities. In-Rack CDUs accommodate up to 250kW of cooling capacity for liquids up to 45°C, with redundant power supplies and pumps as standard for maximum reliability.
Rear Door Heat Exchanger
Rear Door Heat Exchangers attach directly to server racks to remove heat at the source, significantly reducing data center air conditioning requirements while maintaining optimal operating temperatures for high-density AI infrastructure.
Cooling Hose Kit
Flexible hose kits designed to connect liquid-cooled racks directly to data center primary water supply or cooling towers. Pre-installed hose kits enable plug-and-play, toolless deployment with seamless integration into new or existing facility piping.
Power Shelves & BBUs
33kW Power Shelves engineered for the latest generation of high-density AI racks, combined with Battery Backup Units providing 90 seconds of 48V DC power for application checkpointing during power failures – eliminating the need for complete application restarts.
Networking Switches
NVIDIA Ethernet and InfiniBand platforms including NVIDIA Spectrum-X and Quantum-2 for optimized AI cluster networking, as well as qualified partner solutions. Supermicro Ethernet switches with up to 800 GbE link speeds and 51.2 Tbps throughput are also available for diverse workload requirements.
In-Row Solutions
Scale your liquid cooling infrastructure with flexible in-row solutions designed for multi-rack deployments. Supermicro In-Row CDUs deliver up to 1.8MW of cooling capacity across multiple racks, while Liquid-to-Air Sidecars provide up to 200kW of standalone cooling with no external infrastructure modifications. For complete turnkey deployment, SuperCluster delivers pre-validated, plug-and-play multi-rack solutions with integrated networking fabric and rigorous L11/L12 testing—enabling seamless scaling from pilot projects to full production.
In-Row CDU
High-capacity In-Row Coolant Distribution Units designed to accommodate multiple high-powered liquid-cooled server racks, delivering up to 1.8MW of cooling capacity for scalable, multi-rack deployments in dense AI data center environments.
Liquid-to-air (L2A) Sidecar CDU
Petascale and object storage servers with seamless integration across leading software-defined networking partners, delivering maximum throughput and scalability for data-intensive AI workloads and large-scale AI model training and high-volume inference.
SuperCluster
Highly customizable, plug-and-play cluster solutions with multiple racks and integrated networking fabric. Each system undergoes rigorous L11 or L12 validation testing before shipping, ensuring turnkey deployments that enable seamless scaling from pilot projects to full production AI infrastructure.
Site Infrastructure Solutions
Build a high-performance data center foundation with fully engineered cabling, power, and cooling infrastructure. Supermicro delivers optimized cabling design and implementation with complete documentation including routing, port mapping, and cable length optimization that reduces time-to-online, material costs, and labor. For site-level thermal management, Supermicro offers water cooling towers in scalable configurations and dry cooling towers that eliminate water consumption – all designed for easy installation and parallel expansion as your infrastructure grows.
Cabling
Optimized traffic engineering with complete cabling design, documentation, and implementation services including routing, port mapping, and cable length optimization. Engineered solutions reduce time-to-online, material costs, and labor while ensuring high-performance data flow for multi-rack deployments.
Water Cooling Towers
Primary liquid loop facility supporting scalable cooling capacity from 1 MW to 50 MW or more with modular designs for faster time-to-operation. Closed-loop systems with EC fan design and N+1 submersible pumps ensure energy-efficient, reliable operation with reduced power and water usage versus conventional air-cooled systems.
Dry Cooler (Coming soon)
Delivers low PUE/WUE cooling for water-constrained regions, with scalable capacities from 1 MW to over 50 MW. Adiabatic-assisted mode enables air pre-cooling for superior performance in high-temperature environments. Closed-loop design keeps coolant clean and non-corrosive, ensuring exceptional efficiency, minimal maintenance, and long-term reliability.
Data Center Transformer (Available in early 2026)
Pad-mount transformers delivering robust, low-impedance power for AI rows and multi-MW blocks. Configurable for MV primaries and secondaries with smart monitoring and loop-feed options, integrating seamlessly with MV switchgear, UPS/BESS, and prefabricated skids for fast, resilient deployments.
Data Center Generators (Available in early 2026)
Standby generators providing fast-start, high step-load acceptance with native paralleling and seamless automatic transfer switch integration. Build scalable N+1/N+N backup architectures from edge rooms to large campuses, ensuring mission-critical uptime for AI compute environments with unified, predictable operations.
Management Software Suite
From bare metal to AI workloads, SuperCloud Software Suite as part of Data Center Building Block Solutions delivers end-to-end management for modern data centers. Unify infrastructure control, automate deployment pipelines, empower developers with self-service AI tools, and optimize multi-tenant GPU cloud operations—all purpose-built for the scale and complexity of AI factories and hyperscale environments.
SuperCloud Composer (SCC)
SuperCloud Composer delivers unified rack-scale and liquid cooling management across servers, networks, PDUs, CDUs, and 3rd-party systems. It provides power management, advanced leak detection, protection, and alerts to safeguard GPU and cooling infrastructure. With observability and tuning at scale, SuperCloud Composer manages 20K+ hosts through a single portal.
SuperCloud Automation Center (SCAC)
SuperCloud Automation Center delivers pre-built, enterprisegrade automation for data center and edge infrastructure. It covers everything from firmware and OS provisioning to Kubernetes and AI workload enablement while ensuring security, scalability and governance.
SuperCloud Developer Experience Console (SDX)
SuperCloud Developer Console is a developer-oriented AI experience console that brings GPU as a Service (GPUaaS), Kubernetes, Machine Learning pipelines, and data services into one workspace. Unlike other infrastructure-only consoles, it delivers turnkey developer agility with self-service provisioning, AI-native workflows, and built in observability, empowering teams to go from code to model faster and safer.
SuperCloud Director (SCD)
SuperCloud Director delivers multi-tenant AI cloud control with integrated bare metal, Ethernet & Infiniband network multi-tenancy, and storage management, optimizing performance, security, and developer agility. Unlike generic cloud management tools, this software is purpose-built for GPU as a Service (GPUaaS), AI factory deployments, and hyperscale operations.
Services and Onsite Deployment
Accelerate your data center deployment with comprehensive professional services from planning through ongoing support. Supermicro Global Services delivers end-to-end expertise including data center design, solution validation, and professional on-site deployment – whether you're building from bare land, retrofitting air-to-liquid cooling, or deploying in co-location facilities. Our integrated approach reduces time-to-online and ensures higher-quality installations, backed by continued on-site support and 4-hour response time options for mission-critical uptime.
Planning and Validation
Projects are thoroughly planned, including designs for the data center floor plan, rack layout, port mapping, and more. Before shipping, Supermicro goes beyond industry standards with testing that includes rack (L11) and cluster-level (L12) validation.
Onsite Deployment
End-to-end deployment starts with Supermicro’s white-glove delivery service, ensuring the utmost care. On-site service team members handle racking and stacking, with optional software installation available.
Onsite Support
Supermicro ensures a smooth transition after deployment, with full documentation and ongoing support to ensure long-term success. Global Services provide options for on-site response time as short as 4 hours and parts replacement services.
Data Center Fit-Out Service
Turn-key solution transforming data center shells to operational environments. Includes design, engineering, permitting, electrical from PDUs to servers, mechanical systems with racks and CDUs, and liquid cooling infrastructure for complete readiness.
Build Service
Buildout services significantly reduce lead time for new data center construction. Supermicro ensures success for bare-land builds, data center retrofitting for liquid-cooling with end-to-end project management.
Contact Supermicro today to design your next-generation AI data center.