Dell, HP, Cisco, Others Unveil Servers Running on New Intel Xeon Chips

Wednesday Feb 19th 2014 by Jeffrey Burt
Share:

The high-end systems are designed to tackle data-intensive workloads like big data, business intelligence and analytics and to challenge RISC systems.

Dell, Hewlett-Packard and Cisco Systems are among the almost two-dozen server makers offering new and enhanced systems that will be powered by Intel's new high-end Xeon E7 v2 processors.

Intel's new 22-nanometer chips, based on the Ivy Bridge architecture and designed to bring twice the performance, three times the memory and four times the I/O capacity, are aimed workloads such as data analytics, business intelligence and high-performance computing, where performance, reliability and a lot of memory are key. The processors offer up to 15 cores and up to 1.5 terabytes of memory per socket, and improved in-memory processing capabilities.

Intel officials said that 21 OEMs—including IBM, Lenovo, Unisys and SGI—will offer more than 40 system designs based on the new chips, which were code-named Ivytown and announced Feb. 18.

For the system makers, while the new chips are an important part of what they're offering, the differentiators come from the technologies OEMs bring to the systems themselves.

"Innovation matters," Jim Ganthier, vice president of global marketing for HP's Servers Group, told eWEEK.

HP on Feb. 18 announced the ProLiant DL580 Generation 8 system, which will be powered by the new Intel Xeon processors and will offer up to 30 times the speed for business transactions as it leverages the in-memory processing capabilities of the chips. Intelligent management features and infrastructure consolidation will improve efficiency and reduce total cost of ownership by 45 percent, while the use of HP Advanced Error Recovery and HP Memory Quarantine features will mean up to a 30 percent improvement in memory and processor reliability over previous systems, according to HP officials.

Memory Quarantine enables the system to set aside memory if fatal errors are found, while HP's Advanced Error Recovery enables the system to recover from fatal errors in the processor, cache and memory during execution.

The upgraded four-socket server also will include a range of embedded and intelligent management features found in HP's Gen8 systems, including the company's ProActive Insight architecture, which analyzes thousands of system points to improve application performance and uptime.

"We're bringing Gen8 features into the enterprise scale of products," Andrew Hawthorn, group manager of enterprise server product management for HP, told eWEEK.

Intel is aiming its new x86 server chips at high-end workloads like big data, analytics, high-performance computing (HPC) and business intelligence, and is looking to challenge the traditional RISC server makers—particularly IBM with its Power architecture and Oracle with its SPARC-based systems—in the highest end of the server space. HP is aiming the ProLiant DL580 Gen8 at enterprises running such workloads as enterprise resource planning (ERP), data warehousing, online transaction processing, databases, and virtualization and consolidation.

The HP server can be ordered now, starting at $13,079.

Dell introduced its high-end PowerEdge R920, which targets similar data-intense workloads—ERP, databases (including large in-memory databases), virtualization and the like, according to Brian Payne, executive director of server solutions at Dell. The 4U (7-inch) four-socket system includes a range of Dell features, including its Fault Resilient Memory that when combined with Intel's Run Sure Technology increases the servers reliability and availability. In addition, the integration of OpenManage Essentials and Dell AppAssure software, and the use of iDRAC with Lifecycle Controller for automated embedded server configuration backup, improves the continuity of the system's hardware stack, operating system and applications.

The server, which will be available later this quarter, will offer 6TB of memory and up to 24 local storage drives, which will enable organizations to more quickly respond to their changing business environments.

"Customers … are looking for larger memory and larger headroom," Payne told eWEEK.

 

He said the R920 also will support Dell's Fluid Cache for SAN application acceleration solution, which CEO Michael Dell demonstrated during the company's Dell World 2013 event in December. The demonstration showed the data center technology delivering more than 5 million IOPS (input/output operations per second) of random reads.

Cisco rolled out three new Unified Computing System (UCS) servers powered by the Xeon E7 v2 processors. The vendor unveiled the C460 M4 rack system—a four-socket, 4U server—and two blade systems, the four-socket B460 M4 and two-socket B260 M4. Cisco officials said the new systems—which like all UCS servers comes with such features as Cisco's UCS Management and SingleConnect technologies—offer organizations even higher-end options within the portfolio.

All this comes at a time when such trends as cloud computing, mobile, big data and data analytics are growing, demanding more processing power and memory capacity. Cisco's B460 M4 system offers up to 6TB of memory and leverages the in-memory processing capabilities of the new Xeons.

"We've talked about memory scaling. … But dominating performance is the name of the game: behemoth processing performance is what we look for at this end of the server spectrum and Intel has not disappointed on this round of new technology," Todd Brannon, director of product marketing for unified computing at Cisco, wrote in a post on the company blog. "The next generation of the Intel Xeon E7 family packs up to 15 cores per processor and delivers an average 2x performance increase compared to previous generation products. Performance will be even higher on specific workloads, for example, up to 3X on database and even higher for virtualization."

Other OEMs announcing systems powered by Xeon E7 v2 chips include NEC, which rolled out the four-socket Express5800/A2000 series scalable enterprise server; SuperMicro, with the four-socket X10 SuperServer; and Fujitsu, with the PrimeQuest 2000 family of high-end servers that offer up to eight sockets, which company officials say offers the performance of Unix systems and the cost efficiencies of x86.

"The server computing market is going through revolutionary times," Hans-Dieter Wysuwa, executive vice president of products technology and channel for Fujitsu's international business unit, said in a statement. "Our new PrimeQuest infrastructure solution presents enterprises with a new choice of platform for running their mission-critical applications, while also providing the perfect building block for incorporating disruptive technologies like in-memory computing."

 

Originally published on eWeek.
Share:
Home
Mobile Site | Full Site
Copyright 2017 © QuinStreet Inc. All Rights Reserved