Virtualizare cu noile procesoare Intel速 Xeon速 5500
Florin Raclariu Market Development Manager Intel Romania 23 Aprilie 2009 1
Top of Mind Questions for any Business or IT Professional in 2009?
• Why Invest now? • If I do … Where should I invest?
2
Why Invest?
3
We’re in a WW Recession …
1 year S&P 500 Performance, Jan 9, 2009
“Job loses stack up as recession deepens” –Jan 9, 2009
Associated Press
… Business Spending is Tight
4
But Internet Continues to Grow ‌
More customers Data Rich Traffic E-Business E-Commerce Social Media Demands On Server Infrastructure Is Not Slowing Down 5
A Collision of Requirements
“Recession or not, executives want IT to improve competitiveness, not just cut costs” – Dec 16, 2008
Information Week
6
Where Should I Invest?
7
How Healthy is Your Data Center? IT is spending too much on maintenance Innovate
Enhance
Maintain Source: Gartner IT Key Metrics Data 2008.
1 Source: Intel. See Legal information slides for detailed comparisons.
8
This limits ability to innovate and grow business with new services and capabilities
"Investing in new infrastructure routinely, like refreshing older servers, is critical. At Intel IT, we found that continually enhancing our existing infrastructure actually reduces the amount required to maintain and run our environment and therefore is a high priority investment for us.� —Diane Bryant, CIO, Intel Corporation
9
Server Workloads 2008 MCS Server Lifecycle Question 5.6:
Approximately how many years do you estimate that your XXOS servers will be kept in operation?
Figure 33 Server Deployment Lifecycle 45%
The data was surprisingly similar by company size and vertical industry 10
Median
40% 35%
Average is ~4yr
30% 25%
Most Typical is 5yr and 3yr
20% 15% 10% 5% 0%
1
2
3
4
5
6 Total (N=1108)
7
8
9
10+
Let’s Look Closer at Some of These Scenarios
11
Server Refresh Benefits (Single Core)
2005
2009
9x
Performance Refresh
Up to Performance
1:1
Intel®
Xeon®
184 5500 Based Servers
18%
Annual Energy Costs Estimated Reduction
– OR – 90%
Efficiency Refresh
Annual Energy Costs Estimated Reduction
9:1
184 Intel® Xeon® Single Core Servers
Intel®
Xeon®
21 5500 Based Servers
As Low as
8 Month Payback
Source: Intel estimates as of Nov 2008. Performance comparison using SPECjbb2005 bops (business operations per second). Results have been estimated based on internal Intel analysis and are provided for informational purposes only. Any difference in system hardware or software design or configuration may affect actual performance. For detailed calculations, configurations and assumptions refer to the legal information slide in backup.
12
Modernizing Your Infrastructure by replacing old SPARC* with Intel Xeon Processors
Old SPARC*
New Intel
V890* (8S/16C)
X4450* (4S/24C)
SUN* UltraSPARC* IV+
Quad-core Intel速 Xeon速 processor 7400
20 Servers
2 Servers
Realize the Annual Savings of replacing SPARC with Intel Hardware
$550k Savings1
Software License
$480k Savings1
Software Support
$485k Savings1
System Mgt
$1.5m Savings1
Energy costs
$60k
Savings1
1) Estimated Savings from Intel RISC TCO Tool. Example. Oracle* Database running on Solaris* migrating to Oracle DB running on Solaris on Xeon For notes and disclaimers, see legal information slide at end of this presentation.
13
Intelligent Investment Decisions in 09
Refresh with Xeon 5500. Pays for itself in ~8 months. Virtualize with Xeon. IT flexibility & cost advantages. Migrating from RISC to Intel reduces TCO, saving $ Using Intel 4S servers for your most data demanding applications Accelerating research, speed development with Intel HPC/WS solutions Expanding services using Intel technology for Cloud Computing Boosting energy efficiency, save $ and support Green IT with Intel Accelerating ROI upgrading HW with new software Deployments Supporting new buyers in emerging markets & SMB with entry Xeon servers
14
Introducing the Intel速 Xeon速 5500 Series Processor
Based on the Next Generation Intel速 Microarchitecture
A A New New Generation Generation of of Intelligent Intelligent Servers Servers
15
Intel® Xeon® Processor 5500 Series (Codename Nehalem-EP)
A tremendous step forward in technology – – – – –
New New New New New
processor architecture platform architecture memory subsystem I/O subsystem options with SSDs
An even bigger step forward for IT capability
Performance
16
Energy Efficiency
Virtualization
Delivering Intelligent Performance Next Generation Intel® Microarchitecture
Bandwidth Intensive • Intel® QuickPath Technology • Integrated Memory Controller
Intel® 5520 Chipset Intel® Node Manager Technology
Intel® Data Center Manager
Threaded Applications
• 45nm quad-core Intel® Xeon® Processors • Intel® Hyper-threading Technology
PCI Express* 2.0
Performance on Demand Intel® X25-E SSDs
ICH 9/10
Intel® 82599 10GbE Controller
• Intel® Turbo Boost Technology • Intel® Intelligent Power Technology
Performance Performance That That Adapts Adapts to to Your Your Software Software Environment Environment All future products, dates, and figures are preliminary and are subject to change without any notice.
17
Performance Summary
Intel速 Xeon速 Processor 5500 Series (Nehalem-EP) High Performance Computing
Technical Compute Servers
Up to
3.5X Finite Element Analysis
Bandwidth
Comp Floating Fluid Point Dynamics
General Purpose
Mainstream Enterprise Servers
Memory Bandwidth
Weather
Up to
2.25X
Performance
Energy
Database
Virtualization
SAP Java
Intel Xeon X5570 (2.93 GHz) vs. Intel Xeon X5482 (3.20GHz)
Integer
Intel Xeon X5570 (2.93 GHz) vs. Intel Xeon X5460 (3.16GHz)
Compelling Compelling Performance Performance Gains Gains Across Across the the Board Board Source: Published/submitted/approved results March 30, 2009. See backup for additional details
18
An Alternative to RISC Up to
70% More Performance
Intel Xeon速 5570 2S/8C vs. SUN UltraSPARC* T2+ 2S/16C Higher is better Java
Up to Java
40% Lower System Cost 50% Fewer Cores
1.19
SAP
1.62
Integer
1.69
Floating Point
1.71
1.22
1.00
Intel Xeon X5570 (2.93GHz) vs. Sun UltraSPARC* T2+ (1.4GHz)
B aseline SP ECjA ppServer2004 SA P -SD (2 Ttier)
SP ECjbb2005
SP EC SP ECo mpM base2001 int_rate_base 2006
Source: UltraSPARCT2+ results published on SPEC.Org. Intel estimates as of Feb 2009. Intel results have been estimated based on internal Intel analysis and are provided for informational purposes only. Any difference in system hardware or software design or configuration may affect actual performance. For detailed calculations, configurations and assumptions refer to the legal information slide in backup. All systems priced with 32GB memory, T5240 actual cost from published pricing information, Estimated price used for 2S Xeon system, based on existing 2S system published pricing information. For more information, see legal information slide at end of this presentation.
19
Intel速 Xeon速 5500 Performance Publications SPECint*_rate_base2006 240 score
(+71%)
SPECjAppServer*2004
(+93%)
SPECpower*_ssj2008 1860 ssj_ops/watt IBM J9* JVM
(+64%)
TPC*-C
SPECfp*_rate_base2006 194 score
(+125%)
SAP-SD* 2-Tier
(+130%)
(+102%)
3,975 JOPS Oracle WebLogic* Server
631,766 tpmC Oracle 11g* database
5,100 SD Users SAP* ERP 6.0/IBM DB2*
SPEComp*Lbase2001
TPC*-E
SPECWeb*2005
(+189%)
(+152%)
(+140%)
234,996score Intel Compiler 11.0/RHEL 5.3*
800 tpsE Microsoft SQL Server* 2008
71,045 score Rock Web* Server
Fluent* 12.0 benchmark
SPECjbb*2005
SPECapc* for Maya 6.5
(+127%)
Geo mean of 6 ANSYS FLUENT*
604,417 BOPS IBM J9* JVM
(+64%)
(+87%)
7.70 score Autodesk* Maya
Over 30 New 2S Server and Workstation World Records! Percentage gains shown are based on comparison to Xeon 5400 series; Performance results based on published/submitted results as of March 30, 2009. Platform configuration details are available at http://www.intel.com/performance/server/xeon/summary.htm *Other names and brands may be claimed as the property of others Performance tests and ratings are measured using specific computer systems and/or components and reflect the approximate performance of Intel products as measured by those tests. Any difference in system hardware 20 or software design or configuration may affect actual performance. Buyers should consult other sources of information to evaluate the performance of systems or components they are considering purchasing. For more information on performance tests and on the performance of Intel products, visit Intel Performance Benchmark Limitations
Energy Efficiency Summary Energy Efficiency
• Up to 2.25x performance, similar power envelope† • Lower system idle power reduces IT costs • Dynamically turns cores on/off to meet performance needs
Idle and Peak Power Reductions†
Matching Power Consumed to Performance Delivered
12% System Power
50%
5300 00 5 5 on
Xeon
Idle
Xe
ea Id
lP
ow
er
i Ut
50% System Utilization
tio a liz
n
Server technology designed to better optimize cost and performance real time
100%
† Source: Intel internal measurements using SPECjbb2005* as of Aug 2007. System configurations: 2S, 80W processors, 8 DIMMs, 1 HDD, 1 PSU. Except Irwindale=110W processors; Power measurements using SPECjbb2005. Tylersburg-EP from Intel internal measurements as of Sept 2008 with 2.93GHz 95W processors. This information is preliminary and subject to change before launch. For more information, see legal information slide at end of this presentation.
21
Intel® Xeon® 5500 Series:
First Platform With End-to-end HW Virtualization
Intel® VT-x
Intel® VT For Directed I/O
Intel® VT For Connectivity
Processor
Chipset
Network
Extended Page Tables
Intel® VT-d
Intel® VT-c
Intel® VT Flex Migration
Intel Platform Virtualization Technologies Intel® VT Flex Priority
22
Virtual Machine Device Queues (VMDq)
Enhancing Virtualization Performance Better Performance†
Lower Overheard††
VMmark – VMware ESX
“Roundtrip” Virtualization Latency
Up to 2.6x Higher Performance
~40% Quicker over Xeon® 5400 Intel QPI 18 slots DDR3 memory
EPT VPID
PCIe2
More VMs
Virtualization Software Efficiency
(Per Server, Per Core, Per Ft2, Per Watt)
(Nehalem Microarchitecture, EPT, VPID)
† Source: Published/submitted/approved results March 30, 2009. See backup for additional details †† Source: Intel internal measurements. Intel Xeon processor 5500 series (Nehalem) vs. Intel® Xeon® processor 5400 series (Penryn) and Intel® Core™ microarchitecture (Merom)
23
Flexible Virtualization
Through Leadership Processor, Chipset, and I/0 Enhancements
Consolidation
24
Live Migration
More Resources, Hardware Assist
Enhanced I/O, Compatible Architecture
CapEx and OpEx Reduction Improved Utilization
Fluid Movement of VMs Over Network
25
Product Options
Intel® Xeon® Processor 5500 Series CPU Number
Usage
Cache
Link Speed
Max Memory Speed
Turbo
†
HT
8M
6.4 GT/s
1333 MHz
Up to 400 MHz
√
4
8M
5.86 GT/s
1066 MHz
Up to 266 MHz
√
2-4
4M
4.8 GT/s
800 MHz
No Turbo
No HT
Cores
95 W
X5570 (2.93 GHz) X5560 (2.80 GHz) X5550 (2.66 GHz) 80W
Advanced
E5540 (2.53 GHz) E5530 (2.40 GHz)
Standard
80W
60 W
E5520 (2.26 GHz) E5506 (2.13 GHz)
L5520
Low
L5506
Power
80W
E5504 (2.00 GHz) E5502 (1.86 GHz)
Basic
4
Dual-core
Value Value Add Add Features Features Available Available on on Higher Higher End End Processors Processors † Max Turbo Boost frequency based on number of 133 MHz increments above base freq (+2 = 0.266 GHz, +3 = 0.400 GHz).
26
Summary Performance
• Up to 3.5x the bandwidth for technical computing† • Up to 2.25x performance for enterprise computing† • 2x software threads and performance boost on demand
Energy Efficiency
• Up to 2.25x performance, similar power envelope† • Lower system idle power reduces IT costs • Dynamically turns cores on/off to meet performance needs
Virtualization
• Expand existing pools with Intel VT FlexMigration • Unique capabilities for live VM migration over network • Highest 2S virtualization scores mean more VMs per server†
® Xeon ® 5500 Refresh Refresh With With Intel Intel® Xeon® 5500 Series: Series: Boost Performance, Lower IT Costs, 8-month Boost Performance, Lower IT Costs, 8-month ROI ROI † Source: Intel internal measurements, January 2009. For notes and disclaimers, see performance and legal information slides at end of this presentation.
27
28