5 Jun 2018 Taipei - Super Micro Computer Inc., a global expert in enterprise computing, storage, networking solutions and green computing technology, has focused on green computing with Resource Saving data centre and Cloud solutions, including SuperBlade, BigTwin, all-flash NVMe composable storage systems and AI and Machine Learning optimized servers at Computex Taipei 2018.
Data centres account for 3% of the global electricity supply and 2% of the total greenhouse gas emissions. In addition, there is a turnover cycle of outdated electronic systems every three to five years which go into our landfills. This e-waste accounts for about 70% of overall heavy metal waste that poses risks to the environment and human health.
With Resource Saving technology, Supermicro has introduced an overall architecture that optimizes data centre power, cooling, shared resources, and refresh cycles. This innovative approach focuses on reusing system enclosures, enabling modular refresh of subsystems and using optimized extended life subsystems including networking, storage, cooling fans and power supplies. By disaggregating CPU and memory, each resource can be refreshed independently, allowing data centres to both reduce refresh cycle costs and optimize the adoption of new and improved technologies.
"Supermicro has deployed Resource Saving Systems in volume at many Fortune 100 data centres", stated Charles Liang, President and CEO of Supermicro. "As the innovative leader in energy efficient systems and green computing technology, Supermicro now enables further savings by extending the life of server subsystems including the chassis enclosure, cables, networking, storage, fans and power supplies. The result is up to 60% hardware acquisition cost savings, up to 50% less power consumption and less e-waste during technology refresh cycles, which saves data centres millions in costs and helps preserve the environment for future generations."
Supermicro will be showcasing a breadth of platforms at Computex that feature Resource Saving technology and address a wide range of workloads.
The new Supermicro disaggregated SuperBlade unlocks the interdependence between the major server subsystems enabling the independent upgrade of CPU + Memory, I/O, Enclosure, Storage and Power/Cooling. Now each component can be refreshed at the optimal time to maximize generational improvements in performance and efficiency uncoupled from single monolithic server refresh cycles. The blade servers feature Intel Xeon Scalable processors and support M.2 NVMe and Intel Optane.
Delivering high performance and efficiency in a 2U 4-node design, the Supermicro BigTwin system also supports the full range of Intel Xeon Scalable processors, fully exploits all the memory channels with a maximum of 24 DIMMs per node, and offers options for up to six hot-swappable all-flash NVMe or hybrid NVMe/SAS3 drive bays, up to three PCI-E 3.0 slots including support for a flexible SIOM module enabling 100/40/25/10/1G networking options and redundant 2600W/2200W Titanium Level (96%+) digital power supplies.
Also, Supermicro's comprehensive portfolio of AI and Machine Learning (ML) systems based on the latest generation CPUs and NVIDIA Tesla V100 with NVLink GPUs for superior performance and density will be featured including the new 2 PetaFLOP SuperServer design based on the NVIDIA HGX-2 that combines 16 Tesla V100 32GB SXM3 GPUs connected via NVLink and NVSwitch to deliver breakthrough AI and ML compute power.
Supermicro Rack Scale Design (RSD) will also be on display. Supermicro RSD is based on Intel RSD, an industry-aligned data centre architecture built on open standards. Supermicro RSD manages racks of disaggregated servers, storage, and networking with industry-standard Redfish Restful APIs that remain consistent across different vendors and multiple server generations. Supermicro RSD 2.1 supports high performance, high density, and disaggregated NVMe storage to improve data centre efficiency, increase utilization and lower costs.
The exhibit will also include the brand new all-flash NVMe Samsung NF1 and Intel "ruler" form factor 1U storage servers and JBOF that support up to 0.5PB of latest NVMe storage. To ensure the highest data bandwidth with lowest latency, both architectures support dual 100G Infiniband, dual 100G Ethernet as well as NVMe over Fabrics thru two PCI-E x16 Gen3 lanes.