The 10 Hottest Enterprise And Business Servers Of 2018

Here are the hottest enterprise and business servers of 2018. Cisco, Dell EMC, HPE, Inspur, IBM and Supermicro are leading the server market innovation pace.

ARTICLE TITLE HERE

The Hottest Servers Of 2018

The server market continued to skyrocket in 2018 as worldwide sales increased at a blistering pace thanks to innovation from leaders like Dell EMC and Hewlett Packard Enterprise.

As enterprise and cloud service providers continue to spend their budget on updating and expanding data centers, vendors are injecting new technologies to boost server performance and capabilities including new processors and artificial intelligence. The overall server market experienced historic demand in 2018, with consecutive quarters of double-digit revenue growth and its highest total revenue in the second quarter at $22.5 billion, according to research firm IDC.

CRN breaks down some of the hottest server technology in the market in 2018.

id
unit-1659132512259
type
Sponsored post

Cisco C480ML

This year, Cisco Systems launched its new artificial intelligence, machine-learning UCS server: the C480ML. Cisco's server, which is Nvidia GPU-loaded, looks to capitalize on the big data boom as the number of customers running big data workloads on UCS servers has increased significantly over the past few years. The C480ML comes with eight Nvidia Tesla V100 data center GPUs, up to 24 drives able to handle 182 TB, up to six NVMe drives and the ability to handle up to four 100-GB network switches. Cisco, San Jose, Calif., is offering validated designs that use the C480ML as a foundation while integrating with software like Google Cloud Platform, Kubernetes, Docker and Apache Spark to handle Hadoop, containers and AI and machine-learning workloads.

Cisco C4200

The networking leader spent a year creating the new UCS C4200 Series Rack Server Chassis and nodes that target compute-intensive, scale-out workloads. The new C125 M5 Rack Server Node has up to 129 percent higher processor core density compared with a 42RU rack of UCS servers. Each node contains 16 DIMM slots for up to 2 TB of capacity and 2,666MHz of DDR4 memory. It contains two AMD Epyc 7000 CPU processors, marking the first time Cisco used AMD and not Intel processors. The nodes are inside Cisco's new C4200 chassis, which is a multinode offering that optimizes application performance and operational efficiency. The C4200 can be managed with all UCS and HyperFlex systems from a single cloud-based console to make it easier to deploy, scale and manage infrastructure.

Dell EMC PowerEdge R940xa

The Round Rock, Texas-based server leader unveiled its four-socket PowerEdge R940xa server this year, designed to accelerate applications to deliver real-time decisions for GPU database workloads. It combines up to four CPUs with four GPUs in a 1:1 ratio and enables low latency with direct-attached NVMe drives. The R940xa has mix-and-match capacity and performance options with up to 32 drives as well as RESTful API for developers. The server was created to tackle different performance bottlenecks with improved response times and reduce statistical errors compared with older PowerEdge generations. It is optimized for workloads that are compute-layer-heavy and leverage CPU RAM to move data into GPU RAM for faster computation.

Dell PowerEdge MX740c

The two-socket blade delivers a fully configurable compute modular server with high performance and a rich set of storage options including NVMe drives. The single-width MX740c supports the full Intel Xeon Scalable Processor family with up to 6 TB of memory. The server holds up to 24 DDR4 DIMMs slots, 3-TB max memory and speeds of 2666MT/s. The MX740c has more effective cooling and disruptive fabric upgrades compared with other generations as well as having no midplane design. The MX740c is one of the building blocks for Dell EMC's composable PowerEdge MX architecture. The PowerEdge MX portfolio is designed with kinetic infrastructure that has built-in future-proofing to easily support configurations down to memory-centric devices such as GPUs and storage-class memory.

HPE ProLiant DL360 Gen9

Hewlett Packard Enterprise's build-to-order ProLiant was the highest-selling server line through distribution in North America in the second quarter of 2018, according to research firm The NPD Group. The DL360 Gen9 model was the top-selling product of the quarter. The HPE ProLiant DL360 Gen9 server delivers a 1U chassis with up to two processors that combines high performance, low energy consumption and increased density compared with previous generations. The server leverages Intel's E5-2600 v4 processors with the latest 2,400MHz HPE DDR4 SmartMemory that supports up to 3 TB. Customers can manage the server in any IT environment by automating server life-cycle management tasks with HPE OneView and HPE Integrated Lights-Out. Palo Alto, Calif.-based HPE said InfoSight—its cloud-based AI management tool—will be available on ProLiant servers in January 2019.

HPE Superdome Flex Server

The HPE Superdome Flex Server was designed to power critical applications that enable real-time analytics and tackle data-intensive high-performance computing workloads. The server has scalability from four sockets all the way up to 32, with a maximum memory of 48 TB. The Superdome Flex Server includes Intel Xeon Scalable processors with speed of up to 3.6GHz. With top-notch RAS and end-to-end security, the server aims to safeguard vital workloads by leveraging technology such as Error Analysis Engine, which predicts hardware faults and initiates self-repair without operator assistance.

IBM Power E950

Armonk, N.Y.-based IBM launched its new four-socket, 4U form factor Power System E950 server this year that is ideal for cloud deployments through built-in virtualization and flexible capacity. Targeting medium and large enterprises, the IBM Power9-based server has built-in security at all layers—from the processor to the operating system. The E950 provides up to 48 Power9 cores and up to 16 TB of DDR4 memory. The system comes with a built-in IBM Power virtual machine and has 10 slots of PCI. Big Blue's E950 has a single system backplane and single cooling domain.

Inspur AI AGX-5

Dubbing it the most powerful AI supercomputer, Inspur, based in Jinun, China, unveiled the AGX-5 server in November equipped with 16 Nvidia Telsa V100 Core CPUs with up to 32 GB. The 8U form factor AGX-5 is capable of computing deep learning workloads at up to 2 petaflops per second within a single server. Based on Nvidia's latest HGX-2 platform, the product leverages Nvidia's NVSwitch interconnect fabric. The server is equipped with two 28-core CPUs that provide general-purpose computing performance and 6 TB of persistent memory for high-speed data access.

Lenovo ThinkSystem SR850

Lenovo's new ThinkSystem SR850 is a four-socket server that features a streamlined 2U rack design optimized for performance, flexibility and scalability. Models of the SR850 are powered by up to four Intel Xeon Processor Scalable Family processors, each with up to 28 cores. The SR850 is designed for rapid upgrades for processors and memory with flexible storage capacity to keep pace with data growth. The server can scale from two to four Intel Xeon Gold or Platinum processors. Built for standard workloads like general business applications and server consolidation, it can also accommodate high-growth areas such as databases and virtualization. The Hong Kong-based company’s server supports up to 48 DIMMs and mix-and-match storage with up to 16 drives.

Supermicro SYS-6049GP-TRT

Supermicro's new 4U SuperServer 6049GP-TRT provides the performance required to accelerate the diverse need of applications around artificial intelligence and the Internet of Things. For maximum GPU density and performance, the server supports up to 20 Nvidia Tesla T4 Tensor Core GPUs, 3 TB of memory, and 24 hot-swappable 3.5 drives. The 6049GP also features four, 2000-watt Titanium-level efficiency redundant power supplies to help optimize power efficiency, uptime and serviceability. The San Jose, Calif.-based company’s product aims to dramatically increase the density of GPU server platforms for wide data center deployments supporting deep learning applications.