Revolutionary Supermicro Universal GPU System
The Universal GPU system architecture combines the latest technologies supporting multiple GPU form factors, optimized CPU choices, storage and networking options together to deliver highly scalable, single-configuration systems. Systems can be optimized for each customer’s specific artificial intelligence (AI), machine learning (ML) and high performance computing (HPC) applications. Organizations around the world are demanding new options for their next generation computing environments that have the thermal headroom for the next generation of CPUs and GPUs.
“Supermicro’s Universal GPU Server is a breakthrough for our industry and our data center infrastructure,” said Charles Liang, President and CEO, Supermicro. “With a single powerful, modular and scalable platform, we’ve changed the ground rules for how CTOs and IT administrators can plan new deployments. It’s a flexible and revolutionary system design that is simply the best available and will enable large-scale deployments.”
Initially, the Universal GPU Platform will support systems containing all 3rd Gen AMD EPYC™ 7003 with MI250 GPUs or the NVIDIA A100 Tensor Core 4-GPU, and the 3rd Generation Intel Xeon Scalable processors with integrated AI accelerators and NVIDIA A100 Tensor Core 4-GPU. These systems are designed with enhanced thermal capacity to accommodate GPUs up to 700W.
The Supermicro Universal GPU Platform is designed to work with a wide range of GPUs based on open standards design. By adhering to an agreed set of hardware design standards, such as Universal Baseboard (UBB) and OCP Accelerator Modules (OAM), as well as PCI-E and platform-specific interfaces, IT administrators can choose the architecture GPU best suited to their needs. HPC or AI workloads. This will meet the high demands of many enterprises and simplify the installation, testing, production and upgrades of GPU solutions. Additionally, IT administrators will easily be able to choose the right combination of CPU and GPU to create the most optimal system for their users.
The 4U or 5U universal GPU server will be available for accelerators using the UBB standard, as well as PCI-E 4.0, and soon PCI-E 5.0. Additionally, 32 DIMM slots and a wide range of storage and networking options are available, which can also be connected using the PCI-E standard. The Supermicro Universal GPU Server can accommodate GPUs using base cards in either SXM or OAM form factors that use ultra-high-speed GPU-to-GPU interconnects such as NVIDIA NVLink or the AMD xGMI Infinity Fabric, or connect GPUs directly via a PCI-E slot. All of today’s major CPU and GPU platforms will be supported, giving customers choices to match their exact workloads.
The server is designed for maximum airflow and accommodates current and future CPUs and GPUs where the highest TDP CPUs and GPUs are required for maximum application performance. Liquid cooling (direct on chip) options are available for the Supermicro Universal GPU Server because CPUs and GPUs require increased cooling solutions. In addition, thanks to its modular design, specific subsystems of this server can be replaced or upgraded, extending the life of the entire system and reducing electronic waste generated by a complete replacement with each new generation. CPU or GPU technology.
Supporting quotes
“The NVIDIA A100 Tensor Core GPU accelerates all AI applications and over 2,700 HPC applications,” said Paresh Kharya, Senior Director of Product Management for Data Center Computing, NVIDIA. “Supermicro’s new server platform integrates the NVIDIA A100 to give businesses a powerful system that can deliver exceptional performance and productivity for a wide range of workloads, including HPC, AI inference, training and machine learning.”
“AMD appreciates our close relationship with Supermicro. We look forward to Supermicro Universal GPU systems, which will be an ideal platform to demonstrate our advanced innovations in AMD EPYC CPUs and AMD Instinct GPUs that meet industry needs. HPC and on-demand for compute-accelerated data center workloads.” – Brad McCredieCorporate Vice President, Data Center GPUs and Accelerated Processing, AMD
Additional information about the Universal GPU is available online as well as a webinar scheduled for April 21, 2022at 10:00 a.m. PT.
About Supermicro
Supermicro (NASDAQ: SMCI) is a global leader in total application-optimized computing solutions. Founded and operating in San Jose, California, Supermicro is committed to delivering first-to-market innovation for enterprise IT infrastructure, cloud, AI and 5G Telco/Edge. We are transforming into a total IT solution provider with systems, software and services of servers, AI, storage, IoT and switches, while offering advanced motherboard, power and high volume chassis. Products are designed and manufactured in-house (in the United States, Taiwanand the Netherlands), leveraging global operations for scale and efficiency and optimized to improve TCO and reduce environmental impact (Green Computing). The award-winning Server Building Block Solutions® portfolio enables customers to optimize their exact workload and application by choosing from a broad family of systems built from our flexible, reusable building blocks that support a comprehensive set of form factors, processors, memory, GPUs, storage, networking, power and cooling solutions (air conditioning, free air cooling or liquid cooling).
Supermicro, Server Building Block Solutions and We Keep IT Green are trademarks and/or registered trademarks of Super Micro Computer, Inc.
All other brands, names and trademarks are the property of their respective owners.
SOURCESuper Micro Computer, Inc.