Data center News
10 Hottest New Enterprise Servers Of 2020
Mark Haranas
From Dell Technologies and HPE to SuperMicro and Lenovo, here are the hottest new enterprise servers of 2020.
10 New Servers Powering The Data Center, Cloud And Enterprise
A slew of new enterprise servers were launched in 2020 backed by new processing power to give modern data centers and enterprises next-generation server capabilities ideal for modern application workloads such as AI and Kubernetes.
From new Nvidia chips to flash storage technologies, new servers launched this year from the likes of market leaders such as Dell Technologies, Lenovo and Hewlett Packard Enterprise, as well as smaller players like Bamboo Systems and Penguin Computing, is revolutionizing the performance and workload abilities a server can provider. Servers continue to be the base on which data centers are built on.
CRN breaks down the ten hottest new enterprise servers launched in 2020.
Get more of CRN’s 2020 tech year in review.
Bamboo Systems B1000N
Bamboo Systems’ new Arm-based B1000N 1U server is based on its Parallel Arm Node Designed Architecture (PANDA). PANDA-based servers use embedded systems methodologies designed to run modern microservices-based workloads while consuming minimal energy and delivering the density needed for high throughput computing.
Bamboo’s B1000N features a modular blade design that allows for flexible swapping of any blade and choice of component. The server has NVMe SSD storage up to 8-TB per compute node and 32-TB per blade. Each B1000N server can be configured with either one or two blades in 1U, with each blade containing four compute nodes and a nonblocking embedded L3 switch exposing dual 40-Gbit QSFP uplink ports.
Dell EMC PowerEdge XE2420
Dell Technologies introduced this year the Dell EMC PowerEdge XE2420, a highly configurable, dual-socket 2U rack server that targets low-latency, large storage edge applications. It features two second-generation Intel Xeon Scalable processors, with further performance coming from up to four Nvidia acceleration cards.
The PowerEdge XE2420 supports up to 100-TB of flash storage using Intel Enterprise and Datacenter SSD Form Factor drives. Software includes Canonical Ubuntu, Microsoft Windows/Hyper-V, Red Hat Enterprise Linux 8, Suse Linux Enterprise Server and VMware ESXi.
For edge computing use cases, the PowerEdge XE2420 can leverage iDCRAC9 (integrated Dell Remote Access Controller 9) with the optional data center capability to provide streaming telemetry with enhanced integrated security for emerging edge applications.
Hewlett Packard Enterprise Superdome Flex 280
HPE introduced the Superdome Flex 280 this summer, a server targeting digital transformation as well as managing SAP HANA and Oracle databases or massive IoT and artificial intelligence data sets at the edge or in the core.
The HPE Superdome Flex 280, built on HPE’s Superdome Flex family of servers, starts at two and scales up to eight 3rd generation Intel Xeon Scalable Processors. With 6 UPI links per processor, the server provides higher bandwidth and faster data rates than prior generations. It is designed to support 64-GB to 24-TB of shared memory and optimized for in-memory data analytics. Customers can choose between all-DRAM systems or a combination of DRAM and Persistent Memory with Intel OptaneTM persistent memory 200 series.
The Superdome Flex 280 is available through HPE’s GreenLake as-a-service offering.
Inspur AI NF5488A5
Launched in May, NF5488A5 is Inspur’s new AI server powered by 8 Nvidia A100 GPUs fully interconnected with the 3rd generation NVLink, as well as two AMD CPUs supporting PCIe 4.0. The NF5488A5 server provides users with ultra-high-speed bandwidth while offering powerful computing support for various AI scenarios such as intelligent customer service, financial analysis, smart cities and intelligent language processing.
Inspur’s NF5488A5 supports up to eight double-width A100 PCIe cards in a 4U chassis. The server can be configured with up to 40 GB of high bandwidth memory. The NF5488A5 is equipped with 2 AMD Rome CPU and offers AI performance of 5 petaFLOPS. It has a high bandwidth of GPU-GPU Communication of 600-GB/s.
Lenovo ThinkSystem SR645
Lenovo launched a new two-socket server platform with the ThinkSystem SR645 tailor-made to handle the explosion of data around critical workloads. The new server utilize up to two AMD EPYC 7002 CPUs and 128 cores with class-leading memory speed, storage and GPU density.
The SR645 is a 1U form factor with Lenovo increasing onboard storage of up to 40 2.5” drives or up to 32 NVMe drives. The boost in storage allows for dense software-defined storage solutions in data centers. Lenovo’s unique AnyBay backplane technology provides storage flexibility and performance, with support for up to 12x 2.5-inch SAS/SATA HDDs/SSDs.
The ThinkSystem SR645 offer 32 DDR4 memory slots and 4TB using 128GB RDIMMs. The company says its high-speed, low latency NVMe storage is ideal for caching data utilized in analytics and database deployments to gain better business insights and take control of the growing volume and variety of data.
Lenovo ThinkSystem SR860 V2
In June, Lenovo unveiled the ThinkSystem SR860 V2 server targeting analytics and artificial intelligence workloads. The server feature third-generation Intel Xeon Scalable processors with enhanced support for SAP HANA based on Intel Optane persistent memory 200 series, and can be configured with up 12TB of system memory, up to 14x PCIe slots, and up to 48x 2.5-inch drive bays.
The Lenovo ThinkSystem SR860 V2 is a 4-socket server that features a 4U rack design with support for four double-wide 300W or eight single-wide GPUs for handling artificial intelligence workloads, virtual desktop infrastructure or data analytics. The Lenovo ThinkSystem SR860 V2 provides the advanced capabilities of four of the new third-generation Intel Xeon Scalable processors plus support for four double-wide GPUs such as the NVIDIA Tesla V100S GPU. This combination provides a significant processing power in one server.
Nor-Tech Voyageur
Voyageur is a brand of powerful, customer-built server from Nor-Tech, a custom-system builder specializing in high-performance computing. This year, Nor-Tech introduced the Voyageur 8 GPU server. The server, available in 2U and 4.5U models, features up to eight Nvidia GeForce RTX 2080 Ti GPUs based on the Nvidia Turing architecture.
The new eight-GPU server comes with two Intel Xeon Processor Scalable CPUs, two PCIe 3.0 half-length low-profile slots, and room for up to 24 DIMMs and eight hot-swappable 2.5-inch storage drives. Nor-Tech’s 8 GPU Server has an option for 1 Gb or 10 Gb ethernet along with 14 Gbps of GDDR6 memory speed, and a large 11 GB frame buffer.
Penguin Tundra AP
Tundra AP, the latest generation of Penguin Computing’s highly dense Tundra supercomputing platform, combines Intel Xeon Scalable 9200 series processors with Penguin’s Relion XO1122eAP server that delivers a high density of CPU cores per rack. Penguin Computing dubs the Tundra AP as a new Open Compute Project (OCP) server that can max the compute density to more than 7,616 CPU cores in a rack.
Tundra AP is a 21-inch 1U server chassis that can fit two of Intel‘s S9200WK nodes, which each housing two Xeon Platinum 9200 processors, for a total of up to 224 cores per server. Key to Tundra AP is the OCP’s disaggregation design, which removes the power supply unit from the rack server and moves the power sources to power shelves and centralized DC busbars, the latter of which are connected to the motherboard through a simplified power distribution board. This means Tundra AP servers save space and have fewer moving parts and easier thermal design compared with standard servers.
SuperMicro AS-2124GQ-NART
The AS-2124GQ-NART is SuperMicro’s new 2U dual-socket GPU server and is part of its A+ server family which comes with four A100 SXM GPUs, thanks to its use of Nvidia‘s HGX A100 compute board. The GPUs are interconnected using Nvidia’s third-generation NVLink and connected to the CPUs through PCIe 4.0 connectivity. Using AMD’s second-generation EPYC lineup, the processors support up to 64 cores each for a total of 128 cores in the server.
In terms of memory, the SuperMicro AS-2124GQ-NART server has 32 DIMM slots for DDR4-3200MHz SDRAM. On the storage side, it comes with four 2.5-inch hot-swap drive bays for SATA, NVMe or SAS storage. It also has four PCIe 4.0 x16 slots and one PCIe 4.0 x8 slots. For networking, the server comes with two 10 GbE-aggregate host LAN ports and a 1 GbE dedicated IPMI management port.
Tyan Thunder HX FT83-B7119
This year, Tyan launched its Thunder HX FT83-B7119, a new 4U server featuring high-density local storage supporting up to 10 Nvidia V100S or 20 T4 GPUs. The server is based on dual-socket second-generation Intel Xeon Scalable processors, and comes with room for 12 hot-swap 3.5-inch drive bays and 11 PCIe slots.
Tyan’s Thunder HX FT83-B7119 supports 100 Gigabit InfiniBand or Ethernet as well as 24 DIMM slots supporting up to 3-TB. The server‘s chassis includes tool-less drive trays to help ease assembly and service. The server’s storage capacity provides either twelve 3.5” SATA 6G bays or eight SATA plus four NVMe bays with a hot-swap drive carrier design for better serviceability.
Learn More:
Servers
