Supermicro Multi-Node Servers: Max Density and Efficiency
Multi-Node Systems Maximize Compute Density While Saving Power and Space
- Supermicro’s SuperBlade utilizes shared, redundant components, including cooling, networking, power and chassis management, to deliver the compute performance of an entire server rack in a much smaller physical footprint.
- Supermicro’s FlexTwin is a liquid cooled dual processor multi-node architecture optimized for maximum performance density at rack scale. With highly customizable storage, networking, and power supply options, FlexTwin is ideal for demanding HPC workloads including financial services, manufacturing, scientific research, and complex modeling.
- Supermicro’s GrandTwin is a versatile and flexible single processor solution for customers seeking to maximize compute density, scalability, and power efficiency in data center environments.
- Supermicro’s BigTwin offers a range of configurations ideal for hyperscale cloud data centers, with high density compute and storage options enabling customers to maximize space utilization and increase efficiency.
Unmatched Density and Efficiency
Increased Density and Improved Energy Efficiency
- Maximize efficiency with advanced liquid cooling, capturing up to 90% of heat generated.
- Achieve high energy efficiency and compute density with designs that reduce fan power by up to 86%.
- Enhanced efficiency with shared power and cooling architecture.
High Performance Computing
- High performance density with multiple CPUs in compact 2U, 6U, and 8U form factors.
- Support for the latest processor architectures from Intel.
- Delivers numerous efficient cores for demanding workloads.
Cost-Optimized Design
- Liquid cooling reduces server cooling component electricity costs by up to 89%.
- Multi-node/blade configurations per rack offer significant space savings.
- Shared infrastructure optimizes cooling, power, and space for better TCO over time.
Future-Proof Design with High Scalability
- Flexible rear and front I/O options; supports high-performance GPUs.
- SuperBlade offers reusable enclosures, switches, midplanes, power supplies, and fans for next-gen blades.
Multi-Node Servers: Max Density and Efficiency Optimized
Highest Density
Multi-Node Server
- SuperBlade®
- 6U: 5 nodes or 10 nodes
8U; 10 nodes or 20 nodes - Dual or single Intel Xeon 6900/6700/6500 series CPU(s)
- Air-Cooled/Liquid-Cooled
Purpose-Built Liquid-Cooled HPC at Rack Scale Solution
- FlexTwin™
- 4 nodes
- Dual Intel® Xeon® 6900 series CPUs
- Liquid-Cooled
Industry-Leading Multi-Node Architectures
- BigTwin®
- 2 or 4 nodes
- Dual Intel® Xeon® 6700/6500 series CPUs
- Air-Cooled/Liquid-Cooled
Multi-Node Architecture Optimized for Single-Processor Performance
- GrandTwin®
- 2 or 4 nodes
- Single Intel® Xeon® 6900/6700/6500 series CPUs
- Air-Cooled
TECHTalk:
GrandTwin® Multi-Node Systems
Supermicro GrandTwin is a versatile and flexible single processor solution for customers seeking to maximize compute density, scalability, and power efficiency in data center environments.
Watch our short video to learn more!
TECHTalk:
FlexTwin Multi-Node Systems
Supermicro FlexTwin is a purpose-built rack-scale HPC multi-node solution not only for raw performance per dollar but also for energy efficiency.
Watch our short video to learn more!
X14 GrandTwin®
Multi-Node Architecture Optimized for Single-Processor Performance
The GrandTwin® architecture purpose-built for singleprocessor performance and maximum memory density in the data center or at the edge. In addition, GrandTwin® is also highly customizable. You can choose either 2U 2-Node or 2U 4-Node configurations. In fact, not only is adding GPUs optional but also the location of I/O is also an option. Most importantly, you only pay for what you need due to modular architecture.
Architecture purpose-built for single-processor performance and maximum memory density in the data center or at the edge. In addition, GrandTwin® is also highly customizable. You can choose either 2U 2-Node or 2U 4-Node configurations. In fact, not only is adding GPUs optional but also the location of I/O is also an option. Most importantly, you only pay for what you need due to modular architecture.
Key Features
- Single processor per node – lower TDP which leads to lower power and cooling costs
- Front or rear I/O
- High-performance GPU supported
- Support for DDR5-6400 with up to 16 DIMMs per node
- and support for MRDIMMs up to 8800MT/s
- Optional support for EDSFF E1.S NVMe drives
- Flexible AIOM networking for future upgrade
X14 BigTwin®
Industry-Leading Multi-Node Architectures
The BigTwin® is a multi-purpose multi-node system with high flexibility and bandwidth. BigTwin® not only provides 2U 2-Node configurations for greater storage and thermal overhead but also offers 2U 4-Node configurations for customers wanting increased density. Indeed, there are flexible storage options, including PCIe 5.0 EDSFF E3.S and U.2 NVMe drives. Additionally, you may select direct-to-chip liquid cooling to minimize your TCO and support the very top bin processors.
Key Markets/Applications
HCI, HPC, CDN, Cloud Computing, Big Data Analytics, Back-up and Recovery, and Scale-out Storage
Key Features
- Highly configurable 2U 4-node and 2U 2-node systems
- Numerous storage options including PCIe 5.0 EDSFF E3.S and U.2 NVMe drives
- Support for DDR5-6400 with up to 16 DIMMs per node and MRDIMMs up to 8000 MT/s
- Flexible networking with up to 400G Ethernet per node
- Optional direct-to-chip liquid cooling
X14 FlexTwin™
Purpose-Built Liquid-Cooled HPC at Rack Scale Solution
FlexTwin™ is a completely new platform designed for maximum performance, density, and serviceability in a liquid cooled multi-node architecture, featuring support for the latest CPU, memory, storage, and cooling technologies.
Purpose-built to support demanding HPC workloads at scale, including financial services, manufacturing, scientific research, and complex modeling, FlexTwin™ is cost optimized for performance per dollar and can be customized to suit specific HPC applications and customer requirements.
Key Markets/Applications
HPC Data Center, Financial Services, Manufacturing, Climate & Weather Modeling, Oil & Gas, and Scientific Research
Key Features
- All-new 2U 4-node liquid cooled architecture
- Front-accessible nodes with front I/O and storage for cold aisle servicing
- Support for DDR5-6400 and 8800MT/s MRDIMMs
- Optional support for EDSFF E1.S NVMe drives
- Flexible networking options
X14 SuperBlade®
A compact 6U and 8U multi-node system designed for maximum density and dense compute performance. SuperBlade® not only provides customizable configurations, including, a wide range of hard drive form factors, but also supports GPU and liquid cooling. It is also resource-saving, since its management modules, PSUs, switch modules, and enclosures are all reuseable.
Key Markets/Applications
AI/ML Inferencing, Hybrid and Private Cloud, Cloud Computing, Big Data Analytics, Financial Services, HPC, CDN, vSAN, and EDA
6U SuperBlade®
AI & HPC Optimized Multi-Node Architecture
Key Features
- 6U enclosure with 10 single-width or 5 double-width servers, sharing power supplies, cooling fans, Chassis Management Modules, and Ethernet switches
- Supports up to 20 CPUs per system
- Support for DDR5-6400 and MRDIMMs up to 8800 MT/s
- Up to 8 E3.S and 2 M.2 NVMe drives per node
- Up to 4 GPUs or network cards per node
- Direct-to-chip liquid cooling option
6U enclosure with 5 double-width nodes
6U enclosure with 10 single-width nodes
8U SuperBlade®
Highest Density Multi-Node Server Solutions
Key Features
- 8U enclosure with 20 single-width or 10 double-width servers, sharing power supplies, cooling fans, CMMs, Ethernet, and InfiniBand switches
- Supports up to 40 CPUs
- Support for DDR5-6400 with up to 16 DIMMs and MRDIMMs up to 8000 MT/s
- Offers a wide range of hard drive form factors, including U.2, U.3, E1.S, and M.2
- Supports high-performance GPU
- Direct liquid cooling option
8U enclosure with 20 single-width nodes
8U enclosure with 10 double-width nodes