Intel Xeon Processor e5-2600 v2 core of modern data center

It companies around the world are working to virtualize and automate data centers in the hope of helping their business achieve higher value and lower costs, delivering new data-driven services faster and more efficiently. Intel (R) Xeon (TM) processor-based servers provide the foundation for this innovation. These servers account for the vast majority of all servers in the current virtualization center and cloud environment, and can support most of the most high-performance workstations.

Up to 35% performance improvements

The Intel Xeon Processor e5-2600 V2 product family has increased by 50% in both the number of cores and cache capacity in the previous generation, while supporting faster memory and providing additional hardware enhancements specifically for virtualization. These innovations will increase the average performance of the various workloads by 35% compared to the previous generation of servers with Intel Xeon processors. +.2 's performance gains are mainly due to the following three key technologies.

• Higher performance to fully meet peak workload requirements. intel® Core Acceleration Technology 2.03 automatically increases server frequency to take full advantage of extended space for performance and cooling. The second-generation Intel Core-frequency acceleration technology delivers higher frequency and intelligence to help you achieve higher performance and efficiency.

• Floating-point operations are up to twice times more performance-enhancing. +,4 intel® Advanced Vector Extended instruction set (intel® ®AVX) provides a new directive that helps significantly improve the performance of +,5 applications that rely on floating-point or vector computing. Recent upgrades have increased 16/32-bit data conversion speeds, helping specific applications improve performance.

• Improve the efficiency of virtualization in a comprehensive way. Advanced Programmable Interrupt Controller virtualization (APICV) promotes the value of intel® virtualization Technology (INTEL®VT) to new heights. The technology reduces the virtual machine exit by 50%, reducing the proportion of virtualization overhead to the total CPU cost to 4%+,6, further improving performance and scalability.

Eliminate network bottlenecks

The Intel Xeon Processor e5-2600 V2 product family dramatically increases the speed of data movement in a high-density virtual environment, which in turn improves server performance and speeds network and storage communication.

• intel® Integrated I/O delivers up to twice times more bandwidth. The PCI Express 3.0 channel has been directly integrated into the processor chip-up to 80 channels per dual-slot server. This technique reduces the I/O latency by 30%, and +,7 twice times the bandwidth compared to the PCIe 2.0 specification based solution. +,8 storage and technical calculations are further optimized to support x16 opaque bridging and elevated peer-to-peer bandwidth.

• Quickly transfer data to the desired location. intel® Data Direct Transfer technology (intel® ®ddio) supports the direct transfer of data from a storage device to the cache. It also increases I/O performance by 2.3 times times +,9, reducing the overhead of memory access.

Enhanced security and compliance

The Intel Xeon Processor e5-2600 v2 product family will help you protect your business more effectively by improving workload isolation, security policy execution, and data encryption by using intel® security technology.

• Better and faster data protection. intel® Data Protection Technology 10 includes the intel® security key-a technology that generates a high quality security key, and the intel® Advanced Encryption Standard new directive (INTEL®AES-NI), which provides hardware-assisted acceleration and improved data encryption. Together with software from leading vendors such as Oracle, Microsoft and McAfee, these technologies can provide a solid foundation for secure online transactions and enterprise-class data protection without sacrificing application performance.

• Provide solid protection for virtual data centers and cloud environments. Build trusted virtual resource pools to maintain security and compliance more effectively in a dynamic, multi-tenant environment. Intel® Platform Protection Technology 11 includes intel® trusted execution technology, which ensures that systems and management programs only start to "known normal" to require password authentication, avoiding tampering with systems and management programs during or before startup. In addition, intel® platform protection technology includes OS Guard, which provides additional protection against malware by blocking the operating system from user mode or the tampered application in the code page.

Reduce energy costs at all levels

The cost of energy and heat dissipation continues to rise. The Intel Xeon Processor e5-2600 V2 product line increases energy efficiency by 40% compared to servers that use the previous generation of Intel Xeon processors. +,12 This series can help you make the most of your energy by optimizing performance and energy consumption, not only for single servers, but also for racks, entire racks, and entire data centers.

• Industry-leading unit server energy efficiency. Compared with the previous generation of transistors, Intel's 22 nm tri-grid transistor will achieve the same performance of the required energy consumption +, 13 reduced by half. Intel (R) Intelligent Energy-saving technology 14 plays a key role in dynamically optimizing performance and energy consumption at all operating points.

• Automatic control of server power consumption. intel® Node Manager enables you to monitor and control server power consumption and set the maximum power limit for each server. You can use this tool to increase rack density, adjust cooling resources based on actual requirements, improve business continuity, and achieve a dynamic balance of resources to accomplish more tasks with less resources.

• Power optimization across data centers. Intel® Data Center Manager can be integrated into an existing management framework to monitor and manage power and cooling for individual or group of servers. With this tool, you can fully understand the power, cooling, and performance of the entire data center and control it effectively.

Optimized Platform Solutions

Intel is committed to continuously delivering higher value to our customers through the design of comprehensive, highly optimized platform solutions.

• Improve performance with intel® ®10 GB Ethernet for demanding workloads while consolidating your data center traffic onto an extensible cloud-ready network.

• Dramatically improve storage performance at a lower cost by equipping each server with intel® cache acceleration Software (INTEL®CAS) and more than one intel® solid-state disk (INTEL®SSDS).

• In addition, integration of Intel GB Ethernet and Intel Solid state disk with Intel®distribution for Apache hadoop* software has increased the performance of large data analysis by nearly 30 times times. +,15

Intel (R) Xeon (TM) processor E5 v2 Product Family overview

Excellent performance, multi-core, multithreading to meet the widest range of applications and environmental requirements

• Support up to 12 cores and 24 threads per road

• Increase density and performance for virtualized data centers and public/private clouds.

Greater cache and faster memory

• Up to MB level three caching facilitates quick access to frequently used data

• Up to 24 DIMMs per dual processor, multiple data-intensive virtual machines supported

• Higher performance at maximum memory speeds (1866 MHz and 1600 MHz) faster than previous generations to fully meet peak workload requirements

• Intel Core-frequency Acceleration technology 2.03 leverages extended space for performance and cooling to improve processor frequency for peak workloads

Improved performance for HPC applications +,5

• intel® Advanced Vector Extended instruction set (INTEL®AVX) increases vector and floating-point computation speed, supports 256-bit vector and accelerated 32/64-bit data conversions

• Improved virtualization efficiency by increasing the performance of floating-point operations by more than twice times

• Advanced Programmable Interrupt Controller virtualization (APICV) improves Intel virtualization technology due to a 50% reduction in virtual machine exit

• Reduce virtualization overhead to 4% per cent of the total CPU overhead, improving performance and scalability +,6

Optimized data center solutions help reduce costs

Industry-leading I/O performance

• intel® Integrated I/O provides up to 80 pcie* channels (per dual server), a maximum latency reduction 30%,+,7 supports the PCIE 3.0 specification, and increases I/O bandwidth by twice times +,8. It also supports atomic operation and enhanced Peer-to-peer bandwidth.

• Intel (r) Data Direct transfer technology increases I/O performance by 2.3 times times with direct "store to cache" Communications 9 first integrated storage and server processor

• Support for critical storage processor features, including x16 non-transparent bridging that can enhance scalability (contrasting x8-opaque bridging) and RAID 5 and 6 acceleration raid without custom ASIC

Stronger, faster data encryption

Intel data protection technologies include:

• Security key (secure key)-Provides high quality security keys

aes-ni-can perform full encryption without sacrificing application response time to ensure a reliable basis for multi-tenant security

Intel platform Protection technologies include:

• Intel trusted execution Technology-helps IT staff build trusted virtual resource pools to provide more robust security and compliance in multi-tenant virtual and cloud environments

OS guard-can effectively prevent attempts to control the platform or execute malicious software elevation attacks

Industry-leading energy efficiency

• Intel's 22 nm tri-grid transistor reduces energy consumption by half compared to previous generations +,13

• intel® Intelligent Energy-saving technology 14 enables dynamic management of CPU and memory energy consumption status to minimize system power overall monitoring and control without impacting performance

• Intel (r) Node Manager helps IT department monitor and control server power consumption

• intel® Datacenter Manager helps IT department to dynamically optimize energy consumption at all levels, including separate servers, racks, entire racks, and entire data centers

Table 2. Intel (R) Xeon (TM) processor E5 v2 Product family specification

The software involved in the performance testing process and its performance can only be optimized under the Intel microprocessor architecture. Tests such as Sysmark and Mobilemark are based on specific computer systems, hardware, software, operating systems, and functions, and any of these elements are changed

Can lead to changes in test results. Refer to other information and performance tests (including performance when used with other products) to conduct a comprehensive assessment of the target product.

1 Intel processor numbers are not used as a measure of performance. Processor markings are used only to differentiate between processors of the same family, and not to differentiate between different series of processors.

2 for "Up to 35% performance boost" source data:

SPECINT*_RATE_BASE2006: Benchmark configuration and performance metrics evaluation score: RH2288 V2: Two intel® xeon™ processors e5-2690 (2.9 GHz, 8 cores, MB three cache, 8.0 gt/s, 135W), 128 GB Memory (16x 8 GB

2rx4 PC3-12800R-11,ECC), GB SAS K RPM, with kernel 2.6.32-220.el6. Compatible's Red hat Enterprise Linux Server 6.2. Compiler version: C/c++:intel C + + Studio XE for Linux 13.0.0.079

Version。 Fortran:intel Fortran Studio XE for Linux 13.0.0.079 Edition.

New configuration and performance metrics Evaluation score: intel® Server Board S2600CP platform: with two intel® xeon™ processors e5-2697 v2 (2.7 GHz, 12 cores, MB three cache, 8.0 gt/s, 130W, C0 stepping), support EIST, support

Core-frequency acceleration, support for Hyper-threading, 128 GB memory (8x16 GB ddr3-1866), Red hat Enterprise Linux Server 6.3. cpu2006-1.2 with Intel compiler IC13.1. Source: Intel tr#1270 (up to June 2013 11

Day). Estimated score: 638.

Server Virtualization Performance Metrics Evaluation: Benchmark configuration and performance metrics Evaluation score: intel® Server Board S2600GL flat: With two intel® xeon™ processors e5-2690 (2.9 GHz, 8 cores, MB three cache, 8.0 gt/s, 135W, C2 step

In), support virtualization technology, support core-frequency acceleration technology, support Hyper-threading, support NUMA, support MLC Spatial prefetcher, support DCU Data line prefetcher,256 GB memory (16x GB ddr3-1333 DR REG ECC), 128 GB SATA

SSD,ESXI4.1U1 (Build 348481). Source: Intel tr#1359 (as of July 24, 2013). Score: 1723 @ 108VMs.

New configuration and performance metrics Evaluation score: intel® Server Board S2600CP platform with two intel® xeon™ processors e5-2697 v2 (2.7 ghz,12 core, MB three cache, 8.0 gt/s,130w,c0 stepping), virtualization enabled

operation, enabling Intel Core frequency Acceleration technology, enabling Intel Hyper-Threading technology, enabling NUMA, enabling MLC space prefetching, enabling DCU data row prefetching, 256 GB memory (16x GB ddr3-1866 DR REG ECC), 128 GB SATA Solid state Drive, ESXi 5.1

(Build 799733). Source: Intel tr#1359 (as of July 24, 2013). Score: 2246 @ 144VMs.

Performance metrics evaluation of enterprise resource planning: Benchmark configuration and performance metrics Evaluation score: intel® Server Board S2600CP platform with two intel® xeon™ processor e5-2690 (2.9 ghz,8 core, MB three cache, 8.0 gt/s,135w,c2

Stepping), enabling Intel Core frequency Acceleration technology, enabling Intel Hyper-Threading technology, enabling NUMA, enabling MLC space prefetching, enabling DCU data row prefetching, 128 GB memory, SLES11-SP1, oracle* Database 10.2.0.4, creator SAP ECC6.0

EHP4. Source: Intel tr#1706 (as of July 18, 2013). Score: 6,847.

New configuration and performance metrics Evaluation score: intel® Server Board S2600CP platform with two intel® xeon™ Processor e5-2697 v2 (2.7 ghz,12 core, MB three cache, 8.0 gt/s,130w,c0 step), enabling core frequency acceleration

operation, enabling Hyper-Threading technology, enabling NUMA, enabling MLC space prefetching, enabling DCU data row prefetching, 128 GB memory (16x8 GB ddr3-1600), SLES11-SP1-GM, Oracle Database 10.2.0.4, Creator SAP ECC6.0 EHP4. Information to

Source: Intel tr#1706 (as of July 18, 2013). Score: 9,444.

Server-side Java Performance Metrics evaluation: Benchmark configuration and performance metrics Evaluation score: intel® Server Board S2600CP platform: two intel® xeon™ processor e5-2690 (2.9 GHz, 8 cores, MB three cache, 8.0 gt/s, 135W, C2 step

IN), 128 GB memory (16x8 GB ddr3-1600), RHEL 6.2.32, Oracle Hotspot 1.7.0_21-b05. Source: Intel tr#1310 (as of July 18, 2013). Score: 40,047 Max-jops.

New configuration and performance metrics Evaluation score: intel® Server Board S2600CP platform: with two intel® xeon™ processors e5-2697 v2 (2.7 GHz, 12 cores, MB three cache, 8.0 gt/s, 130W, C0 stepping), GB of memory (8x8 GB

ddr3-1866), RHEL 6.2.32, Oracle Hotspot 1.7.0_40-ea-b2. Source: Intel tr#1310 (as of July 18, 2013). Score: 58,279 Max-jops.

3 requires the system to adopt the intel® core-frequency acceleration technology. Specify only Intel (r) processors support intel® Core acceleration technology and intel® core Acceleration Technology 2.0. Please consult your computer manufacturer for details. The final performance of this technology will depend on your hardware, software, and system configuration.

4 performance comparisons based on Linpack performance metrics. Benchmark score 159.4 is based on Intel internal measurements valid prior to December 5, 2011 and is measured using supermicro* X8DTN + dual intel® xeon™ processor X5690, GB

X86_6 Red hat Enterprise Linux Server 6.1 beta system, enabling core-frequency acceleration technology, EIST, and Hyper-Threading technology. New score 347.7 based on Intel internal measurement results, the measurements were made using two intel® xeon™ processors

e5-2690, GB memory, x86_6 Red hat Enterprise Linux Server 6.1 Beta Intel ®rose City platform, enable or disable the core-frequency acceleration technology, enable EIST and Hyper-Threading technology.

5 AVX is designed to enhance throughput for certain integers and floating-point operations. Because of the processor power and heat dissipation characteristics, as well as system power and cooling, AVX floating-point directives may run at a lower frequency to maintain reliable operations in all practices.

6 Configuration: Pre-production 1S intel® xeon™ processor e5-2600 (3.0 GHz, 4 core, MB three cache, B0 stepping), Hyper-threading on, core-acceleration shutdown, EIST shutdown, CPU C-state shutdown, power off, virtualization technology open, NUMA

Open, prefetch function off, the GB solid state disk, xen-unstable c/s23237 w/apic-v patch. Management program without APICV CPU utilization: 5.5%; equipped with APICV Management Program CPU utilization: 4%. Source: Up to

Intel internal measurement results for December 2011.

7 Source: Intel Internal measurement data, the average time required for the I/O device to read data from the local system memory in idle state. Compared with the intel® xeon™ processor e5-2600 product family (230 nanoseconds) compared to the intel® xeon™ Processor 5500 series

(340 nanoseconds) of Ascension.

Benchmark configuration: Green City system: Two intel® xeon™ processor E5520 (2.26 GHz, 4 cores), GB memory @ 1333, disabling C status, disabling core-frequency acceleration technology, disabling smt,rubicon* pcie* 2.0 x8. New configuration: Meridian Department

Two intel® xeon™ processor e5-2665 (C0 Stepper, 2.4 GHz, 8 cores), GB memory @1600 MHz, enable C-State Heripin acceleration technology. The measurement results are based on the Lecroy* pcie* Protocol Analyzer using the Intel internal Rubicon (pcie* 2.0) and Florin (pcie* 3.0) test cards, and the operating environment is Windows * 2008 R2.

The 8 gt/s and 128B/130B encoding features in the 8 PCIe 3.0 specification can increase the interconnect bandwidth to twice times the Pcie* 2.0 specification.

9 Data Source: I/O performance increase to 2.3 times times the declaration based on Intel internal measurement data-Compare intel® xeon™ processor E5 product family and intel® xeon™ processor 5600 series: Two-level forwarding test using 8x10 GBE port

SNB data.

10 no computer system can provide absolute security in all operating environments. You need to install compatible intel® processors and software and optimize for the technology you are using. Please consult your system manufacturer and/or software manufacturer for more information.

11 No computer system can provide absolute security in all operating environments. This technology requires that the technical service be ordered to a qualified service provider and that a compatible chipset, firmware, software (not applicable to all countries) be installed. resulting in any data and/or

Intel will not be liable for loss, theft or other loss of the system. Please consult your service provider for availability and functionality.

12 benchmark configuration and scoring based on server-side Java Energy efficiency performance metrics: intel® Server Board S2600CP platform with two intel® xeon™ processor e5-2690 (2.9 ghz,8 core, MB three cache, 8.0 gt/s,135w), hard

Pre-fetching: disabling; MLC space prefetching: disabling; CU Streamer: disabling; DCU IP: enabled; Balanced performance mode, GB memory (8 x 4 GB ddr3l-1600 ECC REG), Microsoft Windows Server 2008 R2 , IBM J9 VM 1.7.0. To

Score: 4,033. Source: Intel TR #1299 (as at August 12, 2013).

New configuration and scoring based on server-side Java Energy efficiency performance metrics: intel® Server Board S2600CP platform with two intel® xeon™ processors e5-2697 v2 2.7 ghz,12 core, MB three cache, 8.0 gt/s,130w,c0

Stepping), hardware prefetching: disabling; MLC space prefetching: Disabled; CU Streamer: Disabled; DCU IP: Enabled; Intel Hyper-Threading Technology: enabled; Intel Core-Frequency Acceleration technology: enable; Balanced performance mode, GB memory (8 x 4 GB ddr3l-1600 ECC

REG), Microsoft Windows Server 2008 R2 SP1, IBM J9 VM 1.7.0. Score: 5,670. Source: Intel TR #1299 (as at August 12, 2013).

13 is compared with the previous generation of D transistors using 32 nm planar transistor technology. Source: Intel internal test.

14 intel® Smart Power technology requires a computer system with processors, chipsets, basic input and output systems (BIOS) that support intel® smart power technology, and operating systems that apply to certain features. Features or other advantages vary according to hardware implementations,

You may need to update the BIOS and/or the operating system. Please contact your system manufacturer for more information.

15 Source: Intel's Terasort performance metrics evaluation in December 2012. Custom settings: mapred.reduce.tasks=100, Mapred.job.reuse.jvm.num.tasks=-1. Cluster configuration: A header node (named nodes, Job Tracker), 10 Work nodes (data node, Task Tracker), Cisco nexus* 5020 GB switch. Performance data is measured using a lometer* (queue depth of 32). Baseline work node: Supermicro sys-1026t-urf1u server with two intel® Xeon (tm) processors X5690 @ 3.47 ghz,48 GB memory, GB 7200 RPM SATA hard drives, intel® Ethernet server Adapters I350-t2,apache hadoop* 1.0.3,red Hat Enterprise linux*6.3,oracle java.