December 14, 2011
Patented HyperCloud technology enables 1333 MT/s memory speeds on future Intel Xeon E5 family based two-processor servers while LRDIMM only enables 1066 MT/s
IRVINE, Calif., Dec. 13 -- Netlist, Inc., a designer and manufacturer of high-performance memory subsystems, today announced results of side-by-side testing of its HyperCloud modules versus LRDIMM (load reduced dual inline memory module). The test results confirmed that HyperCloud is able to achieve 1333 MT/s (mega transfers per second) memory speeds while LRDIMM is only able to achieve 1066 MT/s.
The memory speed benefit is an important enabling factor for increased application performance for the many enterprises that engage in memory-intensive high performance computing (HPC) simulations, securities transactions and analytics for memory databases. This speed advantage will also be a key contributing factor in the decisions by major OEM server manufacturers to adopt HyperCloud for future generation of servers.
The testing was performed on identical systems at 3 DIMMs per channel (DPC) populated on future Intel Xeon E5 family based two-processor servers. HyperCloud is the patented high-performance virtual dual rank (2vR) RDIMM which overcomes traditional memory bottlenecks that limit server memory capacity and bandwidth.
"The fundamental difference between HyperCloud and LRDIMM is that HyperCloud is based on a distributed buffer architecture that reduces long data path delays by utilizing multiple small buffers located along the bottom (connector) edge of the module reducing the lengths of data signal paths between the data buffer and DRAMs," stated Dr. Hyun Lee, Chief Technology Officer, Netlist. "The LRDIMM, in contrast, uses a single buffer architecture that has longer data signal trace lengths between the DRAM and connector because all the data signals have to pass through a single buffer (register) and back out to DRAM."
Additional information on Netlist's Virtual Dual Rank HyperCloud RDIMMs can be found at http://www.netlist.com/hypercloud.
Netlist, Inc. designs and manufactures high-performance, logic-based memory subsystems for server and storage applications for cloud computing. Netlist's flagship products include HyperCloud, a patented memory technology that breaks traditional memory barriers, NVvault family of products that enables data retention during power interruption, EXPRESSvault, a PCI Express backup/recovery solution for cache data protection and a robust portfolio of high performance and specialty memory subsystems including HyperStream, VLP (very low profile) DIMMs and Planar-X RDIMMs.
Netlist develops technology solutions for customer applications in which high-speed, high-capacity, small form factor and heat dissipation are key requirements for system memory. These customers include OEMs that design and build tower servers, rack-mounted servers, blade servers, high-performance computing clusters, engineering workstations and telecommunications equipment. Founded in 2000, Netlist is headquartered in Irvine, CA with manufacturing facilities in Suzhou, People's Republic of China. Learn more at www.netlist.com.
10/30/2013 | Cray, DDN, Mellanox, NetApp, ScaleMP, Supermicro, Xyratex | Creating data is easy… the challenge is getting it to the right place to make use of it. This paper discusses fresh solutions that can directly increase I/O efficiency, and the applications of these solutions to current, and new technology infrastructures.
10/01/2013 | IBM | A new trend is developing in the HPC space that is also affecting enterprise computing productivity with the arrival of “ultra-dense” hyper-scale servers.
Ken Claffey, SVP and General Manager at Xyratex, presents ClusterStor at the Vendor Showdown at ISC13 in Leipzig, Germany.
Join HPCwire Editor Nicole Hemsoth and Dr. David Bader from Georgia Tech as they take center stage on opening night at Atlanta's first Big Data Kick Off Week, filmed in front of a live audience. Nicole and David look at the evolution of HPC, today's big data challenges, discuss real world solutions, and reveal their predictions. Exactly what does the future holds for HPC?