May 21, 2022

watchever-group

Inspired by Technology

Supermicro Breakthrough Universal GPU System – Supports All Major CPU, GPU, and Fabric Architectures

5 min read


Simplifies Customer Deployments, Provides Supreme Modularity and Customization Selections for AI, ML, and HPC  

SAN JOSE, Calif., March 21, 2022 /PRNewswire/ — Super Micro Laptop, Inc. (SMCI), a world wide chief in enterprise computing, storage, networking methods, and environmentally friendly computing technological know-how, has introduced a groundbreaking know-how that simplifies huge scale GPU deployments and is a foreseeable future evidence design that supports nonetheless to be introduced technologies. The Universal GPU server gives the top adaptability in a resource-preserving server.

The Universal GPU technique architecture brings together the most up-to-date systems supporting multiple GPU variety components, CPU decisions, storage, and networking possibilities optimized alongside one another to provide uniquely-configured and really scalable methods. Programs can be optimized for every customer’s unique Artificial Intelligence (AI), Equipment Learning (ML), and Higher-Functionality Computing (HPC) apps. Businesses around the globe are demanding new selections for their following technology of computing environments, which have the thermal headroom for the next generation of CPUs and GPUs.

“Supermicro’s Common GPU Server is a breakthrough for our business and information centre infrastructure,” said Charles Liang, president, and CEO, Supermicro. “With a one highly effective, modular, and long term-evidence system, we have improved the floor regulations for how CTOs and IT administrators can prepare for new deployments. This is a adaptable and revolutionary procedure design and style that is basically the greatest readily available and will enable fast deployments at scale.”

To begin with, the Common GPU platform will aid units that include the 3rd Gen AMD EPYC™ 7003 processors with possibly the MI250 GPUs or the NVIDIA A100 Tensor Main 4-GPU, and the 3rd Gen Intel Xeon Scalable processors with developed-in AI accelerators and the NVIDIA A100 Tensor Main 4-GPU. These units are intended with an enhanced thermal capability to accommodate up to 700W GPUs.

The Supermicro Common GPU platform is made to do the job with a vast variety of GPUs primarily based on an open benchmarks layout. By adhering to an agreed on established of hardware design requirements, these as Common Baseboard (UBB) and OCP Accelerator Modules (OAM), as very well as PCI-E and platform-precise interfaces, IT directors can choose the GPU architecture most effective suited for their HPC or AI workloads. This will meet up with the demanding needs of lots of enterprises and will simplify the set up, testing, manufacturing, and updates of GPU options. In addition, IT administrators will be capable to simply opt for the suitable blend of CPUs and GPUs to produce the most ideal method for their customers.

The 4U or 5U Universal GPU server will be offered for accelerators that use the UBB standard, as well as PCI-E 4., and shortly PCI-E 5.. In addition, 32 DIMM slots and a wide range of storage and networking choices are obtainable, which can also be related applying the PCI-E conventional. The Supermicro Universal GPU server can accommodate GPUs utilizing baseboards in the SXM or OAM kind aspects that utilize very high-velocity GPU-to-GPU interconnects these as NVIDIA NVLink or the AMD xGMI Infinity cloth, or immediately connect GPUs through a PCI-E slot. All significant existing CPU and GPU platforms will be supported, giving clients choices that match their exact workloads.

The server is created for most airflow and accommodates current and upcoming CPUs and GPUs where by the best TDP CPUs and GPUs are expected for highest software functionality. Liquid cooling possibilities (direct to chip) are readily available for the Supermicro Universal GPU server as CPUs and GPUs have to have increased cooling alternatives. In addition, with its modular style and design, specific subsystems of this server can be changed or upgraded, extending the service existence of the all round system and cutting down the E-squander produced by entire substitute with just about every new CPU or GPU technological innovation technology.

Supporting Offers

“The NVIDIA A100 Tensor Core GPU accelerates all AI applications and above 2,700 HPC purposes,” said  Paresh Kharya, senior director of product or service administration for facts middle computing, NVIDIA. “Supermicro’s new server platform integrates the NVIDIA A100 to offer you enterprises a powerful method that can deliver exceptional functionality and productivity for a extensive array of workloads such as HPC, AI inferencing, instruction, and machine understanding.”

“AMD values our close relationship with Supermicro. We are eagerly looking ahead to the Supermicro Universal GPU methods, which will be an great platform to demonstrate our sophisticated AMD EPYC CPU and AMD Instinct GPU improvements that address the requires of the HPC market and the demand from customers for compute-accelerated knowledge center workloads.” – Brad McCredie, corporate vice president, Info Heart GPU and Accelerated Processing, AMD

Further facts on the Common GPU can be discovered on the net as properly as a webinar scheduled for April 21, 2022, at 10:00 am PT.     

About Supermicro

Supermicro (NASDAQ: SMCI) is a international chief in Application-Optimized Whole IT Answers. Launched and functioning in San Jose, California, Supermicro is fully commited to offering first to industry innovation for Enterprise, Cloud, AI, and 5G Telco/Edge IT Infrastructure. We are transforming into a Whole IT Alternatives supplier with server, AI, storage, IoT, and switch devices, computer software, and expert services whilst providing innovative superior-volume motherboard, power, and chassis products. The products are made and manufactured in-residence (in the US, Taiwan, and the Netherlands), leveraging world functions for scale and efficiency and optimized to improve TCO and minimize environmental impact (Inexperienced Computing). The award-profitable portfolio of Server Creating Block Solutions® will allow clients to optimize for their exact workload and application by picking from a broad spouse and children of units designed from our flexible and reusable building blocks that support a in depth established of sort variables, processors, memory, GPUs, storage, networking, electric power and cooling alternatives (air-conditioned, cost-free air cooling or liquid cooling).

Supermicro, Server Creating Block Options, and We Preserve IT Green are logos and/or registered logos of Super Micro Computer system, Inc.

All other manufacturers, names, and emblems are the residence of their respective entrepreneurs.

(PRNewsfoto/Super Micro Computer, Inc.)

Cision Look at unique content material to down load multimedia:https://www.prnewswire.com/information-releases/supermicro-breakthrough-common-gpu-system–supports-all-big-cpu-gpu-and-material-architectures-301506291.html

Source Tremendous Micro Personal computer, Inc.



Supply link

watchever-group.com All rights reserved. | Newsphere by AF themes.