BestPerf: Performance & Best Practices

This week, Sun continues to highlight the record-breaking performance of its latest update to the chip multi-threaded (CMT) Sun SPARC Enterprise server family running Solaris.  Some of these benchmarks leverage the use of a variety of Sun's unique technologies including ZFS, SSD, various Storage Products and many more. These benchmarks were blogged about by various members or our team and the URLs are shown below.

Messages

• Sun's CMT is the most powerful CPU regardless of architectural/implementation details (#transistors, #cores, threads, MHz, etc.)!
• Performance tests show that Sun can outperform IBM Power6 by more than 2x on a variety of benchmarks.
• Performance tests show Sun's new 1.6GHz CMT systems can be 20% faster than Sun's previous generation 1.4GHz processors, given Sun's continual advancements in both hardware and software.

Benchmark Results Recently Blogged

Sun T5440 Oracle BI EE World Record Performance
http://blogs.sun.com/BestPerf/entry/sun_t5440_oracle_bi_ee

Sun T5440 World Record SAP-SD 4-Processor Two-tier SAP ERP 6.0 EP 4 (Unicode), Beats IBM POWER6 (note1)
http://blogs.sun.com/BestPerf/entry/sun_t5440_world_record_sap

Zeus ZXTM Traffic Manager World Record on Sun T5240
http://blogs.sun.com/BestPerf/entry/top_performance_on_sun_sparc

Sun T5440 SPECjbb2005, Sun 1.6GHz T2 Plus chip is 2.3x IBM 4.7GHz POWER6 chip
http://blogs.sun.com/BestPerf/entry/sun_t5440_specjbb2005_beats_ibm

New SPECjAppServer2004 Performance on the Sun SPARC Enterprise T5440
http://blogs.sun.com/BestPerf/entry/new_specjappserver2004_performance_on_sun

1.6 GHz SPEC CPU2006: World Record 4-chip system, Rate Benchmarks, Beats IBM POWER6
http://blogs.sun.com/BestPerf/entry/1_6_ghz_spec_cpu2006

Sun Blade T6320 World Record 1-chip SPECjbb2005 performance, Sun 1.6GHz T2 Plus chip is 2.6x IBM 4.7GHz POWER6 chip
http://blogs.sun.com/BestPerf/entry/new_specjbb2005_performance_on_the

Comparison Table

Benchmark	Sun CMT	Tier	Software	Key Messages
Oracle BI EE	Sun T5440	Appl, Database	Oracle 11g, Oracle BIEE, ZFS, Solaris	World Record: T5440 Achieved 28,000 users Reference
SAP-SD 2-Tier	Sun T5440	Appl, Database	SAP ECC 6.0,EP4 Oracle 10g, Solaris	World Record 4-socket: T5440 T5440 Beats 4-socket IBM 550 5GHz Power6 by 26% (note1) T5440 Beats HP DL585 G6 4-socket Opteron (note1) Unicode version
SPECjAppServer 2004	Sun T5440	Appl, Database	Oracle WebLogic, Oracle 11g, JDK 1.6.0_14, Solaris	World Record Single System (Appl Tier): T5440 T5440 is 6.4x faster of IBM Power 570 4.7GHz Power6 T5440 is 73% faster than HP DL 580 G5 Xeon 6C Oracle Fusion Middleware
Sun T5440 SPECjbb2005	Sun T5440	Appl	Java HotSpot, OpenSolaris	1.6GHz US T2 Plus CPU is 2.3x faster of IBM 4.7GHz Power6 CPU 1.6GHz US T2 Plus CPU is 21% faster than previous generation 1.4GHz US T2 Plus CPU Sun T5440 has 2.3x better power/perf than the IBM 570 (8 4.7GHz Power6)
Sun Blade T6320 SPECjbb2005	Sun T6320	Appl	Java HotSpot, OpenSolaris	World Record 1-socket: T6320 1.6GHz US T2 Plus CPU is 2.6x faster than IBM 4.7GHz Power6 CPU T6320 is 3% faster than Fujitsu 3.16GHz Xeon QC
SPEC CPU2006	Sun T5440, Sun T5240, Sun T5220, Sun T5120, Sun T6320	all tiers	Sun Studio12, Solaris, ZFS	World Record 4-socket: T5440 1.6GHz US T2 Plus CPU is 2.6x faster than IBM 4.7GHz Power6 CPU T6320 is 3% faster than Fujitsu 3.16GHz Xeon QC
Zeus ZXTM Traffic Manager	Sun T5240	Web	Zeus ZXTM v5.1r1, Solaris	World Record: T5240 T5240 Beats f5 BIG-IP VIPRON by 34%; 2.6x better $/perf T5240 Beats f5 BIG-IP 8800 by 91%; 2.7x better $/perf⁞ T5240 Beats Citrix 12000 by 2.2x; 3.3x better $/perf No IBM result

Virtualization

Sun's announcement also included updated virtualization software (LDOMs 1.1). Downloads are available to existing SPARC Enterprise server customers at: http://www.sun.com/servers/coolthreads/ldoms/index.jsp.  Also look the the blog posting "LDoms for Dummies" at http://blogs.sun.com/PierreReynes/entry/ldoms_for_dummies

Try & Buy Program

Sun is also offering free 60-day trials on Sun CMT servers with with a very popular Try and Buy program: http://www.sun.com/tryandbuy.

Benchmark Performance Disclosure Statements (the URLs listed above go into more detail on each of these benchmarks)

Note1: 4-processor world record on the 2-tier SAP SD Standard Application Benchmark with 4720 SD User, as of July 23, 2009, IBM System 550 (4 processors, 8 cores, 16 threads) 3,752 SAP SD Users, 4x 5 GHz Power6, 64 GB memory, DB2 9.5, AIX 6.1, Cert# 2009023. T5440 beats HP new 4-socket Opteron Servers (HPDL585 G6 with 4665 SD User and HP BL685c G6 with 4422 SD User)

Two-tier SAP Sales and Distribution (SD) standard SAP ERP 6.0 2005/EP4 (Unicode) application benchmarks as of 07/21/09: Sun SPARC Enterprise T5440 Server (4 processors, 32 cores, 256 threads) 4,720 SAP SD Users, 4x 1.6 GHz UltraSPARC T2 Plus, 256 GB memory, Oracle10g, Solaris10, Cert# 2009026. HP ProLiant DL585 G6 (4 processors, 24 cores, 24 threads) 4,665 SAP SD Users, 4x 2.8 GHz AMD Opteron Processor 8439 SE, 64 GB memory, SQL Server 2008, Windows Server 2008 Enterprise Edition, Cert# 2009025. HP ProLiant BL685c G6 (4 processors, 24 cores, 24 threads) 4,422 SAP SD Users, 4x 2.6 GHz AMD Opteron Processor 8435, 64 GB memory, SQL Server 2008, Windows Server 2008 Enterprise Edition, Cert# 2009021. IBM System 550 (4 processors, 8 cores, 16 threads) 3,752 SAP SD Users, 4x 5 GHz Power6, 64 GB memory, DB2 9.5, AIX 6.1, Cert# 2009023. HP ProLiant DL585 G5 (4 processors, 16 cores, 16 threads) 3,430 SAP SD Users, 4x 3.1 GHz AMD Opteron Processor 8393 SE, 64 GB memory, SQL Server 2008, Windows Server 2008 Enterprise Edition, Cert# 2009008. HP ProLiant BL685 G6 (4 processors, 16 cores, 16 threads) 3,118 SAP SD Users, 4x 2.9 GHz AMD Opteron Processor 8389, 64 GB memory, SQL Server 2008, Windows Server 2008 Enterprise Edition, Cert# 2009007. NEC Express5800 (4 processors, 24 cores, 24 threads) 2,957 SAP SD Users, 4x 2.66 GHz Intel Xeon Processor X7460, 64 GB memory, SQL Server 2008, Windows Server 2008 Enterprise Edition, Cert# 2009018. Dell PowerEdge M905 (4 processors, 16 cores, 16 threads) 2,129 SAP SD Users, 4x 2.7 GHz AMD Opteron Processor 8384, 96 GB memory, SQL Server 2005, Windows Server 2003 Enterprise Edition, Cert# 2009017. Sun Fire X4600M2 (8 processors, 32 cores, 32 threads) 7,825 SAP SD Users, 8x 2.7 GHz AMD Opteron 8384, 128 GB memory, MaxDB 7.6, Solaris 10, Cert# 2008070. IBM System x3650 M2 (2 Processors, 8 Cores, 16 Threads) 5,100 SAP SD users,2x 2.93 Ghz Intel Xeon X5570, DB2 9.5, Windows Server 2003 Enterprise Edition, Cert# 2008079. HP ProLiant DL380 G6 (2 processors, 8 cores, 16 threads) 4,995 SAP SD Users, 2x 2.93 GHz Intel Xeon x5570, 48 GB memory, SQL Server 2005, Windows Server 2003 Enterprise Edition, Cert# 2008071. SAP, R/3, reg TM of SAP AG in Germany and other countries. More info www.sap.com/benchmark.

Oracle Business Intelligence Enterprise Edition benchmark, see http://www.oracle.com/solutions/business_intelligence/resource-library-whitepapers.html for more. Results as of 7/20/09.

Zeus is TM of Zeus Technology Limited. Results as of 7/21/2009 on http://www.zeus.com/news/press_articles/zeus-price-performance-press-release.html?gclid=CLn4jLuuk5cCFQsQagod7gTkJA.

SPEC, SPECint, SPECfp reg tm of Standard Performance Evaluation Corporation. Competitive results from www.spec.org as of 16 July 2009. Sun's new results quoted on this page have been submitted to SPEC. Sun Blade T6320 89.2 SPECint_rate_base2006, 96.7 SPECint_rate2006, 64.1 SPECfp_rate_base2006, 68.5 SPECfp_rate2006; Sun SPARC Enterprise T5220/T5120 89.1 SPECint_rate_base2006, 97.0 SPECint_rate2006, 64.1 SPECfp_rate_base2006, 68.5 SPECfp_rate2006; Sun SPARC Enterprise T5240 172 SPECint_rate_base2006, 183 SPECint_rate2006, 124 SPECfp_rate_base2006, 133 SPECfp_rate2006; Sun SPARC Enterprise T5440 338 SPECint_rate_base2006, 360 SPECint_rate2006, 254 SPECfp_rate_base2006, 270 SPECfp_rate2006; Sun Blade T6320 76.4 SPECint_rate_base2006, 85.5 SPECint_rate2006, 58.1 SPECfp_rate_base2006, 62.3 SPECfp_rate2006; Sun SPARC Enterprise T5220/T5120 76.2 SPECint_rate_base2006, 83.9 SPECint_rate2006, 57.9 SPECfp_rate_base2006, 62.3 SPECfp_rate2006; Sun SPARC Enterprise T5240 142 SPECint_rate_base2006, 157 SPECint_rate2006, 111 SPECfp_rate_base2006, 119 SPECfp_rate2006; Sun SPARC Enterprise T5440 270 SPECint_rate_base2006, 301 SPECint_rate2006, 212 SPECfp_rate_base2006, 230 SPECfp_rate2006; IBM p 570 53.2 SPECint_rate_base2006, 60.9 SPECint_rate2006, 51.5 SPECfp_rate_base2006, 58.0 SPECfp_rate2006; IBM Power 520 102 SPECint_rate_base2006, 124 SPECint_rate2006, 88.7 SPECfp_rate_base2006, 105 SPECfp_rate2006; IBM Power 550 215 SPECint_rate_base2006, 263 SPECint_rate2006, 188 SPECfp_rate_base2006, 222 SPECfp_rate2006; HP Integrity BL870c 114 SPECint_rate_base2006; HP Integrity rx7640 87.4 SPECfp_rate_base2006, 90.8 SPECfp_rate2006.

SPEC, SPECjbb reg tm of Standard Performance Evaluation Corporation. Results as of 7/17/2009 on http://www.spec.org. SPECjbb2005, Sun Blade T6320 229576 SPECjbb2005 bops, 28697 SPECjbb2005 bops/JVM; IBM p 570 88089 SPECjbb2005 bops, 88089 SPECjbb2005 bops/JVM; Fujitsu TX100 223691 SPECjbb2005 bops, 111846 SPECjbb2005 bops/JVM; IBM x3350 194256 SPECjbb2005 bops, 97128 SPECjbb2005 bops/JVM; Sun SPARC Enterprise T5120 192055 SPECjbb2005 bops, 24007 SPECjbb2005 bops/JVM.

SPECjAppServer2004, Sun SPARC Enterprise T5440 (4 chips, 32 cores) 7661.16 SPECjAppServer2004 JOPS@Standard; HP DL580 G5 (4 chips, 24 cores) 4410.07 SPECjAppServer2004 JOPS@Standard; HP DL580 G5 (4 chips, 16 cores) 3339.94 SPECjAppServer2004 JOPS@Standard; Two Dell PowerEdge 2950 (4 chips, 16 cores) 4794.33 SPECjAppServer2004 JOPS@Standard; Dell PowerEdge R610 (2 chips, 8 cores) 3975.13 SPECjAppServer2004 JOPS@Standard; Two Dell PowerEdge R610 (4 chips, 16 cores) 7311.50 SPECjAppServer2004 JOPS@Standard; IBM Power 570 (2 chips, 4 cores) 1197.51 SPECjAppServer2004 JOPS@Standard; SPEC, SPECjAppServer reg tm of Standard Performance Evaluation Corporation. Results from http://www.spec.org as of 7/20/09.

SPECjbb2005 Sun SPARC Enterprise T5440 (4 chips, 32 cores) 841380 SPECjbb2005 bops, 26293 SPECjbb2005 bops/JVM. Results submitted to SPEC. HP DL585 G5 (4 chips, 24 cores) 937207 SPECjbb2005 bops, 234302 SPECjbb2005 bops/JVM. IBM Power 570 (8 chips, 16 cores) 798752 SPECjbb2005 bops, 99844 SPECjbb2005 bops/JVM. Sun SPARC Enterprise T5440 (4 chips, 32 cores) 692736 SPECjbb2005 bops, 21648 SPECjbb2005 bops/JVM. SPEC, SPECjbb reg tm of Standard Performance Evaluation Corporation. Results from www.spec.org as of 7/20/09.

IBM p 570 8P 4.7GHz (4 building blocks) power specifications calculated as 80% of maximum input power reported 7/8/09 in “Facts and Features Report”: ftp://ftp.software.ibm.com/common/ssi/pm/br/n/psb01628usen/PSB01628USEN.PDF

One Sun SPARC Enterprise T5440 server with four UltraSPARC T2 Plus processors at 1.6GHz delivered a single-system World Record result of 7661.16 SPECjAppServer2004 JOPS@Standard using Oracle WebLogic Server, a component of Oracle Fusion Middleware, together with Oracle Database 11g.

• This benchmark used the Oracle WebLogic 10.3.1 Application Server and Oracle Database 11g Enterprise Edition. This benchmark result proves that the Sun SPARC Enterprise T5440 server using the UltraSPARC T2 Plus processor performs as an outstanding J2EE application server as well as an Oracle 11g OLTP database server.
• The Sun SPARC Enterprise T5440 server (four 1.6 GHz UltraSPARC T2 Plus chips) running as the application server delivered 6.4X better performance than the best published single application server result from the IBM p 570 system based on the 4.7 GHz POWER6 processor.

• The Sun SPARC Enterprise T5440 server (four 1.6 GHz UltraSPARC T2 Plus chips) demonstrated 73% better performance than the HP DL580 G5 result of 4410.07 SPECjAppServer2004 JOPS@Standard, which used four 2.66 GHz Intel 6-core Xeon processors.
• The Sun SPARC Enterprise T5440 server (four 1.6 GHz UltraSPARC T2 Plus chips) demonstrated 2.3X better performance than the HP DL580 G5 result of 3339.94 SPECjAppServer2004 JOPS@Standard, which used four 2.93 GHz Intel 4-core Xeon processors.
• One Sun SPARC Enterprise T5440 server (four 1.6 GHz UltraSPARC T2 Plus chips) demonstrated 1.9X better performance than the Dell PowerEdge R610 result of 3975.13 SPECjAppServer2004 JOPS@Standard, which used two 2.93 GHz Intel 4-core Xeon processors.
• One Sun SPARC Enterprise T5440 server (four 1.6 GHz UltraSPARC T2 Plus chips) demonstrated 5% better performance than the Dell PowerEdge R610 result of 7311.50 SPECjAppServer2004 JOPS@Standard, which used two Dell R610 systems each with two 2.93 GHz Intel 4-core Xeon processors.
• These results were obtained using Sun Java SE 6 Update 14 Performance Release on the Sun SPARC Enterprise T5440 server and running the Solaris 10 5/09 Operating Environment.
• The Sun SPARC Enterprise T5440 server used Solaris Containers technology to consolidate 7 Oracle Weblogic application server instances to achieve this result.
• Oracle Fusion Middleware provides a family of complete, integrated, hot pluggable and best-of-breed products known for enabling enterprise customers to create and run agile and intelligent business applications. Oracle WebLogic Server’s on-going, record-setting Java application server performance demonstrates why so many customers rely on Oracle Fusion Middleware as their foundation for innovation.

Performance Landscape

SPECjAppServer2004 Performance Chart as of 7/20/2009. Complete benchmark results may be found at the SPEC benchmark website http://www.spec.org. SPECjAppServer2004 JOPS@Standard (bigger is better)

Vendor	SPECjApp-Server2004 JOPS@Standard	J2EE Server	DB Server
Sun	7661.16	1x Sun SPARC Enterprise T5440 32 cores, 4 chips, 1.6 GHz US-T2 Plus Oracle WebLogic 10.3.1	1x Sun SPARC Enterprise T5440 32 cores, 4 chips, 1.4 GHz US-T2 Plus Oracle 11g DB 11.1.0.7
Dell	7311.50	2x PowerEdge R610 16 cores, 4 chips, 2.93 Xeon X5570 Oracle WebLogic 10.3	1x PowerEdge R900 24 cores, 4 chips @ 2.66 Xeon X7460 Oracle 11g DB 11.1.0.7
Sun	6334.86	1x Sun SPARC Enterprise T5440 32 cores, 4 chips, 1.4 GHz US-T2 Plus Oracle WebLogic 10.3	1x Sun SPARC Enterprise T5440 32 cores, 4 chips, 1.4 GHz US-T2 Plus Oracle 11g DB 11.1.0.7
Dell	4794.33	2x PowerEdge 2950 16 cores, 4 chips, 3.3 Xeon X5470 Oracle WebLogic 10.3	1x PowerEdge R900 24 cores, 4 chips, 2.66 Xeon X7460 Oracle 11g DB 11.1.0.6
HP	4410.07	1x ProLiant DL580 G5 24 cores, 4 chips, 2.66 GHz Xeon X7460 Oracle WebLogic 10.3	1x ProLiant DL580 G5 24 cores, 4 chips, 2.66 GHz Xeon X7460 Oracle 11g DB 11.1.0.6
HP	3975.13	1x Dell PowerEdge R610 8 cores, 2 chips, 2.93 GHz Xeon X5570 Oracle WebLogic 10.3	1x PowerEdge R900 24 cores, 4 chips, 2.66 GHz Xeon X7460 Oracle 11g DB 11.1.0.7
IBM	1197.51	1x IBM System p 570 4 cores, 2 chips, 4.7 GHz POWER6 WebSphere Application Server V6.1	1x IBM p5 550 4 cores, 2 chips, 2.1 GHz POWER5+ IBM DB2 Universal Database 9.1

Results and Configuration Summary

Application Server:

Sun SPARC Enterprise T5440

4 x 1.6 GHz 8-core UltraSPARC T2 Plus
256 GB memory
2 x 10GbE XAUI NIC
2 x 32GB SATA SSD
Solaris 10 5/09
Solaris Containers
Oracle WebLogic 10.3.1 Application Server - Standard Edition
JDK 1.6.0_14 Performance Release

Database Server:

Sun SPARC Enterprise T5440

4 x 1.4 GHz 8-core UltraSPARC T2 Plus
256 GB memory
6 x Sun StorageTek 2540 FC Array
4 x Sun StorageTek 2501 FC Expansion Array
Solaris 10 5/09
Oracle Database Enterprise Edition Release 11.1.0.7

Benchmark Description

SPECjAppServer2004 (Java Application Server) is a multi-tier benchmark for measuring the performance of Java 2 Enterprise Edition (J2EE) technology-based application servers. SPECjAppServer2004 is an end-to-end application which exercises all major J2EE technologies implemented by compliant application servers as follows:

• The web container, including servlets and JSPs
• The EJB container
• EJB2.0 Container Managed Persistence
• JMS and Message Driven Beans
• Transaction management
• Database connectivity

Moreover, SPECjAppServer2004 also heavily exercises all parts of the underlying infrastructure that make up the application environment, including hardware, JVM software, database software, JDBC drivers, and the system network. The primary metric of the SPECjAppServer2004 benchmark is jAppServer Operations Per Second (JOPS) which is calculated by adding the metrics of the Dealership Management Application in the Dealer Domain and the Manufacturing Application in the Manufacturing Domain. There is NO price/performance metric in this benchmark.

Key Points and Best Practices

• 7 Oracle WebLogic server instances on the Sun SPARC Enterprise T5440 server were hosted in separate Solaris Containers to demonstrate consolidation of multiple application servers.
• Each appserver container was bound to a separate processor set each containing 4 cores. This was done to improve performance by reducing memory access latency using the physical memory closest to the processors. The default set was used for network & disk interrupt handling.
• The Oracle WebLogic application servers were executed in the FX scheduling class to improve performance by reducing the frequency of context switches.
• The Oracle database processes were run in 4 processor sets using the psrset utility and executed in the FX scheduling class. This was done to improve perfomance by reducing memory access latency and reducing conext switches.
• Oracle Log Writer process run in a separate processor set containing 1 core and run in the RT scheduling class. This was done to insure that the Log Writer had the most efficient use of cpu resources.
• Enhancements to the JVM had a major impact on performance.
• The Sun SPARC Enterprise T5440 used 2x 10GbE NICs for network traffic from the driver systems.

See Also

• SPECjAppServer2004 Results Page

Disclosure Statement

SPECjAppServer2004, Sun SPARC Enterprise T5440 (4 chips, 32 cores) 7661.16 SPECjAppServer2004 JOPS@Standard; HP DL580 G5 (4 chips, 24 cores) 4410.07 SPECjAppServer2004 JOPS@Standard; HP DL580 G5 (4 chips, 16 cores) 3339.94 SPECjAppServer2004 JOPS@Standard; Two Dell PowerEdge 2950 (4 chips, 16 cores) 4794.33 SPECjAppServer2004 JOPS@Standard; Dell PowerEdge R610 (2 chips, 8 cores) 3975.13 SPECjAppServer2004 JOPS@Standard; Two Dell PowerEdge R610 (4 chips, 16 cores) 7311.50 SPECjAppServer2004 JOPS@Standard; IBM p570 (2 chips, 4 cores) 1197.51 SPECjAppServer2004 JOPS@Standard; SPEC, SPECjAppServer reg tm of Standard Performance Evaluation Corporation. Results from http://www.spec.org as of 7/20/09

Last Friday, Chris posted SPECpower results on our new group blog BestPerf called "Interpreting Sun's SPECpower_ssj2008 Publications", it is well worth a read.

In the coming days you will see a wide variety of other results posted, these may include: speccpu, specfp, specint, specjappserver, specweb, specjbb, specomp, specpower, specjvm, specmail, speccpu, igen, Ansys, Nastran, sap-sd, Siebel, peoplesoft, TPC-C, TPC-E, TPC-H, etc.