Supermicro is well known in the high-performance computing (HPC) world for its dense, efficient servers that can pack a lot of computing power into a small footprint. But the company is now looking to make a more significant impact in HPC by expanding beyond just hardware and into software and services.
Supermicro has been at the forefront of several new initiatives to help bolster its position in HPC. These include recent developments in their partnership with H2O.ai to provide optimized water cooling solutions for AI applications and expanded partnerships with NVIDIA and AMD. Supermicro is also launching a new line of GPU-optimized servers based on the NVIDIA T4 platform, and it has unveiled a new software offering aimed at AI and machine learning workloads.
Supermicro has a long history of developing innovative and award-winning products that have set standards in the HPC industry. For example, Supermicro X10 generation motherboards can support 4x NVIDIA GPU computing acceleration technology for lightning-fast scientific applications to deliver groundbreaking performance.
Innovative storage features are also offered with support for up to 1TB solid-state drive (SSD) cache tiering on select MPX models and 8x 2.5″ hot-swap HDD bays per node, along with optional NVM Express (NVMe) high-speed onboard SSDs with enterprise-class reliability and 12Gb/s SAS expandability.
Supermicro servers offer a wide range of computing solutions such as:
• HyperConverged Infrastructure optimized with VMware vSAN
• SuperStorage optimized with NVM Express and all-flash storage
• Data Management Solutions optimized for AI, analytics and HPC workloads: SuperWorkstations – FN22G/FN24R/ FN31G
• HyperBlade – 2U 100% GPU uncompromised density for remote visualization and virtual workstation solutions (WW4526)
• High-density Compute Solution (HDS47) using UCS mini chassis
• JBOD tower solution – 4U 42x 3.5″ hot-swap HDD bays (JBOD-6038R)
Supermicro’s vision is to provide the most comprehensive, highest quality, and most innovative line of HPC solutions to meet the needs of customers. They are committed to driving innovation in this market, and they look forward to working with customers to provide the best possible experience.
Supermicro has been developing products in HPC environments for many years now. Still, it’s only recently started to make a more significant push in attracting more high-profile customers from this space. In one notable deal, Twitter placed a massive order for 280 Supermicro servers to support its data analytics platform.
The company hopes to build on this success with its new initiatives announced this week. The partnership with H2O.ai will see Supermicro teaming up with the AI startup to provide optimized water cooling solutions for AI applications. This will include the integration of H2O.ai’s open-source machine learning platform, Driverless AI, with Supermicro’s hardware.
In addition, Supermicro is expanding its partnership with NVIDIA to include the new T4 platform. This platform is based on NVIDIA’s Volta architecture and offers superior performance for deep learning and inference applications. Supermicro has already developed several servers based on the T4 platform, including single-socket and 4U systems.
Finally, Supermicro unveils a new software offering aimed at AI and machine learning workloads. This will be a vital component of its new Cluster Management Utility or CMU. This is designed to simplify the creation and management of HPC clusters while also delivering comprehensive monitoring capabilities.
These latest initiatives from Supermicro reflect the company’s desire to become a more significant player in high-performance computing applications. In doing so, it can offer a one-stop-shop for all of the hardware, software, and services required for AI and machine learning.
