eWEEK content and product recommendations are editorially independent. We may make money when you click on links to our partners. Learn More.
Dell EMC is turning to Advanced Micro Devices and its Epyc processors to power three new servers that company officials are targeting at the network edge and at high-performance computing (HPC) workloads.
The new systems, announced this week, grows Dell EMC’s 14th-generation portfolio and gives enterprises another processor option beyond Intel. They also reinforce the vendor’s efforts to deliver compute capabilities to the network edge, which officials said is important given the rise of the internet of things (IoT) and the need to bring processing, storage and analytics closer to where those devices are.
Two of the new rack systems, the 1U (1.75-inch) PowerEdge R6415 and the 2U (3.5-inch) R7415 (pictured), are single-socket servers and are designed for edge computing and software-designed storage (SDS) deployments to enable VMware vSAN deployments at the edge, according to officials.
Trying to Get Back Marketshare from Intel
For AMD, the partnership with Dell EMC is the latest step in its efforts to increase its presence in the data center and compete more closely with Intel, which owns more than 95 percent of the server chip market. AMD’s Epyc 7000 chip is based on the company’s “Zen” architecture and is built in a system-on-a-chip (SoC) design.
The processor offers up to 32 cores and 64 threads, up to 4TB of memory, support for 12 to 24 direct NVMe SSD drives, and 128 lanes of PCI Express interconnects. It also supports dense GPU and field-programmable gate array (FPGA) accelerator capabilities in the PowerEdge servers and memory bandwidth of up to 342 GB/s.
Such features in Epyc enable Dell EMC to offer single-socket servers that can run the type of enterprise workloads that before needed two-socket systems. The PowerEdge servers also can scale up or out, and because they can handle more demanding workloads, Dell EMC can offer customers an attractive total-cost-of-ownership (TCO) story, according to Forrest Norrod, senior vice president and general manager of AMD’s Datacenter and Embedded Solutions Business Group.
The Epyc-based systems “excel in a wide range of workloads, including dense virtualization, hybrid-cloud applications, software-defined storage, CAD/CAM, and other memory and I/O dependent applications,” Norrod said in a statement.
Designed for New Workloads
The new servers, which are available now, were developed to handle the new workloads that are emerging in a fast-changing data center environment that is being roiled by such trends as the cloud, IoT and data analytics. The PowerEdge R6415 is an ultra-dense system that can scale out and offers flexible storage capabilities through up to 10 PCI NVMe drives.
Computing is being pushed to the network edge due to the tens of billions of devices, systems and sensors that make up the IoT and are generating massive amounts of data that needs to be collected, storage, processed and analyzed in near real-time. To enable that, enterprises want to move compute and storage out to the edge and have the bulk of the data processing, storage and analytics happen there rather than sending all the data back to the data center.
“As customers deploy more IoT solutions, they need highly capable and flexible compute at the edge to turn data into real-time insights,” Ashley Gorakhpurwalla, president of server and infrastructure systems at Dell EMC, said in a statement.
The R7415 is certified as a vSAN Ready Node, offering up to 20 percent better TCO per four-node cluster for vSAN deployments at the edge than competitive systems, according to Dell EMC officials. The server offers 128 PCIe lanes, 2TB of memory and up to 24NVMe drives.
The PowerEdge R7425 is designed to support containers, hypervisors, virtual machines and cloud computing and run such HPC workloads like computational fluid dynamics. It includes up to 64 cores and dense GPU and FPGA capabilities.
Competition in Server Chip Market Growing
The competition in the server chip market continues to grow as organizations push for a viable second supply to Intel to help drive innovation and price reductions. Intel continues to improve the performance and power efficiency of its chips and grow the capabilities through the use of such technologies as FPGAs, but other vendors are looking to chip away at Intel’s market share.
AMD has made a good impression with Epyc, such as both Microsoft and Baidu using the chip in their cloud environments. At the same time, IBM is pushing its Power architecture through internal development and the OpenPower Foundation, and Arm and some chip partners—including Qualcomm—are pushing for the low-power architecture to find its way into the data center.
According to IDC analysts, new processors from Intel and AMD are helping to drive new server sales in the enterprise. Hyperscale data center operators like Google, Amazon and Facebook had been keeping the global server market afloat for several quarters, but enterprises in the third quarter 2017 began buying systems again after waiting for the new processors, according to the analysts.
Revenue worldwide in the quarter jumped 19.9 percent over the same period in 2016, while server shipments increased 11.1 percent.