Greater than 15 Product Households of Totally Upgraded Servers Optimized for Most Efficiency or Potency, That includes Then-Pace GPU Help, Upper-Bandwidth Reminiscence, 400GbE Networking, E1.S and E3.S Drives, and Direct-to-Chip Liquid Cooling Help, and are In accordance with the Untouched Intel® Xeon® 6900 Line Processors with P-Cores and Intel Xeon 6700 Line Processors with E-cores
SAN JOSE, Calif., Sept. 24, 2024 /PRNewswire/ — Supermicro, Inc. (NASDAQ: SMCI) a Overall IT Resolution Supplier for AI/ML, HPC, Cloud, Deposit, and 5G/Edge, as of late provides fresh most functionality GPU, multi-node, and rackmount methods to the X14 portfolio, which can be according to the Intel Xeon 6900 Line Processors with P-Cores (previously codenamed Granite Rapids-AP). The fresh industry-leading number of workload-optimized servers addresses the desires of contemporary information facilities, enterprises, and repair suppliers. Becoming a member of the efficiency-optimized X14 servers leveraging the Xeon 6700 Line Processors with E-cores introduced in June 2024, as of late’s additions carry most compute density and tool to the Supermicro X14 lineup to assemble the {industry}’s broadest dimension of optimized servers supporting all kinds of workloads from difficult AI, HPC, media, and virtualization to energy-efficient edge, scale-out cloud-native, and microservices packages.
“Supermicro X14 systems have been completely re-engineered to support the latest technologies including next-generation CPUs, GPUs, highest bandwidth and lowest latency with MRDIMMs, PCIe 5.0, and EDSFF E1.S and E3.S storage,” stated Charles Liang, president and CEO of Supermicro. “Not only can we now offer more than 15 families, but we can also use these designs to create customized solutions with complete rack integration services and our in-house developed liquid cooling solutions.”
Licensed shoppers can download early get admission to to finish, full-production methods by way of Supermicro’s Early Send Program or for far off trying out with Supermicro JumpStart.
Those fresh Supermicro X14 methods trait totally re-designed architectures, together with all-new 10U and multi-node method components that help next-generation GPUs and better CPU core densities, up to date reminiscence slot configurations with 12 reminiscence channels consistent with CPU and fresh MRDIMMs, which lend as much as 37% higher reminiscence bandwidth in comparison to DDR5 DIMMs.
The fresh Supermicro X14 crowd contains a lot of fresh methods —a number of of which can be totally fresh architectures—in 3 distinct, workload-specific sections:
- GPU-optimized platforms designed for natural functionality and enhanced thermal capability to help the up-to-date generation and highest-wattage GPUs. Device architectures are constructed from the field up for large-scale AI coaching, LLMs, generative AI, 3-D media, and virtualization packages.
- Prime compute-density multi-nodes together with all-new FlexTwin™,SuperBlade®, and GrandTwin®, which leverage sources corresponding to shared energy and cooling to extend talent, in addition to direct-to-chip liquid cooling on particular fashions to maximise density with out compromising on functionality.
- Marketplace-proven Supermicro Hyper rackmounts mix unmarried or twin socket architectures with versatile I/O and store configurations in conventional method components to support enterprises and information facilities scale up and out as their workloads evolve.
Supermicro’s fresh max-performance X14 methods help the fresh Intel Xeon 6900 form processors with P-cores, future X14 efficiency-optimized methods help the Intel Xeon 6700 form processors with E-cores, which introduced in June 2024. Those methods will even deal socket compatibility with the then Intel Xeon 6900 form processors with E-cores and Intel Xeon 6700 form processors with P-cores in Q1’25 to lend an extra area of flexibleness when optimizing methods for both performance-per-core or performance-per-watt.
“For the first time, Intel is offering two distinct, workload-optimized families of Xeon processors in the same generation, each designed to deliver specific performance and efficiency profiles that can shorten time-to-results, revolutionize compute, power and rack density, maximizing ROI for modern data centers,” stated Ryan Tabrah, VP and GM of Xeon Merchandise at Intel. “With the new additions, Supermicro will be able to provide their customers with even more choice with the introduction of these new performance optimized CPUs for AI and compute-intensive workloads.”
When configured with Intel Xeon 6900 form processors with P-cores, Supermicro methods help fresh FP16 directions at the integrated Intel® AMX accelerator to additional strengthen AI workload functionality. Those methods come with 12 reminiscence channels consistent with CPU with help for each DDR5-6400 and MRDIMMs as much as 8800MT/s, CXL 2.0, and trait extra intensive help for high-density, industry-standard EDSFF E1.S and E3.S NVMe drives.
Supermicro Liquid Cooling Answers
Complementing this expanded X14 product portfolio are Supermicro’s rack-scale integration and liquid cooling features. With an industry-leading world production capability, intensive rack-scale integration and trying out amenities, and a complete suite of control device answers, Supermicro designs, builds, checks, validates, and delivers whole information heart answers at any scale in an issue of weeks.
Supermicro trade in an entire in-house advanced liquid cooling answer together with chilly plates for CPUs, GPUs, reminiscence, Cooling Distribution Gadgets, Cooling Distribution Manifolds, hoses, connectors, and cooling towers. Liquid cooling is definitely integrated in rack-level integrations to extend gadget talent, let go circumstances of thermal throttling, and lowers each the TCO and Overall Value to the Order (TCE) of knowledge heart deployments.
Untouched Supermicro max-performance X14 methods that includes Intel Xeon 6900 form processors with P-cores come with:
GPU-optimized – The best possible functionality Supermicro X14 methods designed for large-scale AI coaching, huge language fashions (LLMs), generative AI and HPC, and supporting 8 of the latest-generation SXM5 and SXM6 GPUs. Those methods are to be had in air-cooled or liquid-cooled configurations.
PCIe GPU – Designed for max GPU flexibility, supporting as much as 10 double-width PCIe 5.0 accelerator playing cards in a thermally-optimized 5U chassis. Those servers are perfect for AI inferencing, media, collaborative design, simulation, cloud gaming, and virtualization workloads.
Intel® Gaudi® 3 AI Accelerators – Supermicro additionally plans to bring the {industry}’s first AI server according to the Intel Gaudi 3 accelerator hosted via Intel Xeon 6 processors. The gadget is predicted to extend talent and decrease the price of large-scale AI type coaching and AI inferencing. The gadget options 8 Intel Gaudi 3 accelerators on an OAM common baseboard, six built-in OSFP ports for cost-effective scale-out networking, and an unengaged platform designed to usefulness a community-based, open-source device stack, requiring refuse device licensing prices.
SuperBlade® – Supermicro’s X14 6U high-performance, density-optimized, and energy-efficient SuperBlade maximizes rack density, with as much as 100 servers and 200 GPUs consistent with rack. Optimized for AI, HPC, and alternative compute-intensive workloads, every node options breeze cooling or direct-to-chip liquid cooling to maximise talent and reach the bottom PUE with the most efficient TCO, in addition to connectivity as much as 4 built-in Ethernet switches with 100G uplinks and entrance I/O supporting a dimension of versatile networking choices as much as 400G InfiniBand or 400G Ethernet consistent with node.
FlexTwin™ – The fresh Supermicro X14 FlexTwin structure is purpose-built for HPC, cost-efficient, and designed to lend most compute energy and density in a multi-node configuration with as much as 24,576 functionality cores in a 48U rack. Optimized for HPC and alternative compute-intensive workloads, every node options direct-to-chip liquid cooling simplest to maximise talent and let go circumstances of CPU thermal throttling, in addition to HPC low latency entrance and rear I/O supporting a dimension of versatile networking choices as much as 400G consistent with node.
Hyper – X14 Hyper is Supermicro’s flagship rackmount platform designed to bring the best possible functionality for difficult AI, HPC, and endeavor packages, with unmarried or twin socket configurations supporting double-width PCIe GPUs for max workload acceleration. Each breeze cooling and direct-to-chip liquid cooling fashions are to be had to facilitate the help of top-bin CPUs with out thermal boundaries and let go information heart cooling prices future additionally expanding talent.
Additionally, now delivery are Supermicro‘s X14 efficiency-optimized methods that includes Intel Xeon 6700 form processors with E-cores:
SuperBlade® – Supermicro’s high-performance, density-optimized, and energy-efficient multi-node platform optimized for AI, Knowledge Analytics, HPC, Cloud, and Undertaking workloads. To be had as a 6U enclosure with 10 or 5 nodes, or an 8U enclosure with 20 or 10 nodes, one rack can trait as much as 34,560 Xeon compute cores.
Hyper – Flagship functionality rackmount servers designed for scale-out cloud workloads, with store & I/O flexibility that gives a customized have compatibility for a large dimension of utility wishes.
CloudDC – All-in-one platform for cloud information facilities, according to the OCP Knowledge Heart Modular {Hardware} Device (DC-MHS) with versatile I/O and store configurations and twin AIOM slots (PCIe 5.0; OCP 3.0 compliant) for max information throughput.
Petascale Deposit – Trade-leading all-flash store density and function with EDSFF E1.S and E3.S drives, permitting unparalleled capability and function in a 1U or 2U chassis.
WIO – Offer versatile I/O configurations in an economical structure to allow optimization of acceleration, store, and networking possible choices to boost up functionality, build up talent and to find the easiest have compatibility for particular endeavor packages.
BigTwin® – 2U 2-Node or 2U 4-Node platform offering stunning density, functionality, and serviceability with twin processors consistent with node and hot-swappable tool-less design. Those methods are perfect for cloud, store, and media workloads with fresh fashions, together with E3.S power help for stunning density and throughput.
GrandTwin® – Objective-built for single-processor functionality and reminiscence density, that includes entrance (chilly aisle) hot-swappable nodes and entrance or rear I/O for more uncomplicated serviceability. Now to be had with E1.S drives for higher store density and throughput.
Hyper-E – Delivers the facility and versatility of our flagship Hyper crowd optimized for deployment in edge environments. Edge-friendly options come with a short-depth chassis and entrance I/O, making Hyper-E appropriate for edge information facilities and telco cupboards. Those short-depth methods help as much as 3 high-performance GPU or FPGA playing cards.
Edge/Telco – Prime-density processing energy in compact method components optimized for telco cupboard and Edge information heart set up. Non-compulsory DC energy configurations and enhanced working temperatures as much as 55° C (131° F).
About Tremendous Micro Pc, Inc.
Supermicro (NASDAQ: SMCI) is a world chief in Software-Optimized Overall IT Answers. Based and working in San Jose, California, Supermicro is dedicated to turning in first to marketplace innovation for Undertaking, Cloud, AI, and 5G Telco/Edge IT Infrastructure. We’re a Overall IT Answers supplier with server, AI, store, IoT, transfer methods, device, and help services and products. Supermicro’s motherboard, energy, and chassis design experience additional allow our building and manufacturing, enabling subsequent technology innovation from cloud to edge for our world shoppers. Our merchandise are designed and manufactured in-house (in america, Taiwan, and the Netherlands), leveraging world operations for scale and talent and optimized to make stronger TCO and let go environmental have an effect on (Inexperienced Computing). The award-winning portfolio of Server Development Cancel Answers® permits shoppers to optimize for his or her actual workload and alertness via settling on from a vast crowd of methods constructed from our versatile and reusable development blocks that help a complete poised of method components, processors, reminiscence, GPUs, store, networking, energy, and cooling answers (air-conditioned, detached breeze cooling or liquid cooling).
Supermicro, Server Development Cancel Answers, and We Stock IT Inexperienced are logos and/or registered logos of Tremendous Micro Pc, Inc.
All alternative manufacturers, names, and logos are the trait in their respective homeowners.
SOURCE Tremendous Micro Pc, Inc.
WANT YOUR COMPANY’S NEWS FEATURED ON PRNEWSWIRE.COM?
440k+
Newsrooms &
Influencers
9k+
Virtual Media
Shops
270k+
Reporters
Opted In



