Intel® Xeon® Processor E7-4800 Product Family - Scalability

Intel Xeon

Intel® Xeon® processor E7-4800 product family

These top-of-the-line processors deliver performance that is ideal for your most data-demanding workloads with improved scalability, and increased memory and I/O capacity, allowing you to handily adapt to changes in short-term business demands, and address requirements for longer-term business growth.

Socket scaling

2- to 4-Socket performance scaling

Consider using fewer servers to lower the total cost of ownership (TCO) with fewer license costs, lower maintenance and cooling, plus smaller datacenter footprint. All driven by the better performance of the next generation Intel® Xeon® processor E7 platforms.

Benchmark description for 2- to 4-socket performance scaling

Across a variety of workloads listed, moving from a 2-socket volume server to the top-of-the-line 4-socket Intel® Xeon® processor E7-4800 product family not only may double your performance, but also brings additional reliability features for mission critical needs. This chart shows several industry benchmarks ran in Intel Labs to highlight the great scalability of the expandable servers.

2- to 4-Socket performance scaling

.

Configuration details Intel® Xeon® processor E7-4800 product family

2- to 4-Socket performance scaling as of 1 April 2011.

Performance comparison across a representative benchmark workload as of 24 February 2011. The comparison presented above is based on two-socket (2S) servers using Intel® Xeon® processor X5690 and four-socket (4S) E7-4870 (codenamed "Westmere-EX") series.

2-socket Intel® Xeon® processor X5690 based platform details (baseline 1.0)

Intel® 5500 Chipset reference platform with two Intel® Xeon® processors X5690 (3.46 GHz, 12MB L3, 6.4 GT/s, 6-core, 130W TDP). Memory and storage varied based on workload.

ERP: 4-socket Intel® Xeon® processor E7-4870 based platform details

Intel® 7500 chipset server system with four Intel® Xeon® processor E7-4870 (30M cache, 2.40GHz, 6.4GT/s Intel® QPI), 256GB memory, SUSE* Enterprise LINUX 11 SP1, SAN CLARiiON CX3-80 RAID0*, SAP ECC 6.0 EHP4. Source: Intel SSG TR#1117.

Integer Throughput: 4-socket Intel® Xeon® processor E7-4870 based platform details

Intel® 7500 chipset server system with four Intel® Xeon® processor E7-4870 (30M cache, 2.40GHz, 6.4GT/s Intel® QPI), 256GB memory, SUSE* Enterprise LINUX 11 SP1, Intel® C++ Compiler XE2011. Source: Intel SSG TR#1131.

Floating-Point Throughput: 4-socket Intel® Xeon® processor E7-4870 based platform details

Intel® 7500 chipset server system with four Intel® Xeon® processor E7-4870 (30M cache, 2.40GHz, 6.4GT/s Intel® QPI), 256GB memory, SUSE* Enterprise LINUX 11 SP1, Intel® C++ Compiler XE2011. Source: Intel SSG TR#1131.

OLTP Warehouse Database: 4-socket Intel® Xeon® processor E7-4870 based platform details

Intel® 7500 chipset server system with four Intel® Xeon® processor E7-4870 (30M cache, 2.40GHz, 6.4GT/s Intel® QPI), 512GB memory, Red Hat* Enterprise LINUX 5.4 w/ 5.5 kernel, 576x X25 SSDs (JBOD) + 144x 36GB FC 15K RPM storage solution, leading database vendor. Source: Intel SSG TR#1120.

Additional information: 1 2

Información de productos y rendimiento

open

1. Las pruebas y clasificaciones de rendimiento se miden utilizando sistemas y/o componentes informáticos específicos y reflejan el rendimiento aproximado de los productos Intel® medidos en dichas pruebas. Cualquier diferencia en la configuración y en el diseño del hardware o del software del sistema puede afectar al rendimiento real. Los usuarios deberían consultar otras fuentes de información para evaluar el rendimiento de los sistemas o de los componentes que consideren adquirir. Si desea obtener más información sobre las pruebas de rendimiento y sobre el rendimiento de los productos Intel, visite www.intel.com/content/www/us/en/benchmarks/resources-benchmark-limitations.html.


2. El rendimiento relativo de cada análisis de rendimiento se calculó tomando el resultado real del análisis para la primera plataforma y asignándole un valor de 1,0 como referencia. El rendimiento relativo de las demás plataformas, objeto de la prueba, se calculó dividiendo el resultado real del análisis para la plataforma de referencia entre cada uno de los resultados de los análisis específicos de cada una de las demás plataformas y, asignándoles un número de rendimiento relativo correspondiente a las mejoras de rendimiento observadas.