Supermicro unveils faster, ‘greener’ server portfolio
Over 50 server models have been unveiled by Supermicro, made up of more than 15 families for Tier 1 server and storage requirements. The X13 server portfolio is based on fourth generation Intel Xeon scalable processors.
The systems based on the new Intel Xeon Scalable processors (formerly codenamed Sapphire Rapids) deliver up to 60 per cent better workload-optimised performance, said Supermicro, with improved security (hardware root of trust and attestation) and manageability. They are also faster with integrated AI, storage, and cloud acceleration. They have high ambient temperature and liquid cooling options to reduce environmental impact and operating costs.
Charles Liang, president and CEO of Supermicro, explained: “In addition to the processor, every major sub-system has been dramatically upgraded with up to two times improvement in memory performance with DDR5, two times better I/O bandwidth with 80 lanes of PCIe Gen 5 I/O that will support higher performance accelerator cards, and 400Gbits per second networking and improved manageability and security.
“The power and cooling capabilities have also expanded to support 350W CPUs and up to 700W GPUs,” he added.
Accelerator engines built into the 4th Gen Intel Xeon Scalable processors, increase efficiency and reduce CPU load for many critical workloads in today’s data centres. These include Intel Advanced Matrix Extensions (Intel AMX) to accelerate deep learning inference and training performance, Intel Data Streaming Accelerator (Intel DSA) to reduce CPU overhead on storage, networking, and data-processing-intensive tasks, Intel QuickAssist Technology (QAT) for hardware offload of popular cryptography and data compression / decompression algorithms, and Intel AVX for vRAN to improve capacity and reduce power consumption for vRAN workloads.
In addition, Supermicro servers will support the new Intel Data Center GPU Max Series (formerly named Ponte Vecchio) across a wide range of servers. The Intel Data Center GPU Max Series contains up to 128 Xe-HPC cores and will accelerate a range of AI, HPC, and visualisation workloads.
The X13 systems have been designed from the ground up with Supermicro’s Building Block Solutions approach delivering maximum performance and efficiency while also providing a high degree of flexibility to suit customers’ specific workloads and scale. Supermicro X13 systems can be powered by one, two, four, or even eight 4th Gen Intel Xeon Scalable processors, each with up to 60 cores and support for next-generation built-in accelerators and GPUs up to 700W from companies including Intel and Nvidia. In addition, the support for a range of open industry standards, including OCP 3.0-compliant Advanced IO Modules (AIOM), EDSFF storage, and OCP Open Accelerator Module (OAM) and SXM GPU interconnect, reduces reliance on proprietary technologies. This enables customers to standardise components across the data centre and incorporate Supermicro systems into existing infrastructure with minimal modifications. Utilising open standards-based server management, Supermicro servers can seamlessly be integrated into existing IT environments.
The X13 portfolio consists of the SuperBlad, GPU servers with PCIe GPUs designed for HPC, AI / ML, rendering and VDI workloads, Universal GPU servers for workloads with the most demanding AI training performance, HPC, and big data analytics, the X13 Hyper series of rackmount servers, the X13 Hyper-E series, designed for edge data centre and telecomms deployments. There is also the X13 BigTwin systems for cloud, storage, and media workloads and GrandTwin which maximises compute, memory and efficiency for workloads such as content delivery networks (CDNs), multi-access edge computing, cloud gaming and high-availability cache clusters and the X13 FatTwin which supports all-hybrid hot-swappable NVMe / SAS / SATA hybrid drive bays for EDA applications.
Other introductions are the SuperEdge, edge servers, CloudDC, WIO, Petascale Storage and MP servers.
