BM Seer Facts & Questions from an Anonymous Sun Source

Sun's even faster SPECweb2005: New Record Sun SPARC Enterprise T5220 UltraSPARC T2

Wednesday Apr 09, 2008

The Sun SPARC Enterprise T5220 obtained a world record SPECweb2005 result 41847 SPECweb2005 with one UltraSPARC T2 running Solaris 10 with a real world web server Sun Java[TM] System Web Server. This is a 13% improvement over the original result published at www.spec.org in Oct '07. Demonstrating Sun Microsystems continued commitment to performance improvements.

The Sun SPARC Enterprise T5220 server delivers 4% greater performance than the four-socket HP ProLiant DL580 G5 with 2.9 GHz Quad-Core Xeon processors. In addition, the Sun SPARC Enterprise T5220 has 1.8X better power-performance and has 3.6X better SWaP.

The Sun SPARC Enterprise T5220 server delivers 41% greater performance than the 2-socket HP ProLiant DL380 G2 with 3.16 GHz Quad-core Intel Xeon 5460 processors. In addition, the Sun SPARC Enterprise T5220 has 1.1X better power-performance and has 1.1X better SWaP.

There are no IBM POWER6 results on the SPECweb benchmark.  The Sun SPARC Enterprise T5220 server is 5.3 times faster than the 4-core IBM p550 1.9GHz POWER5+.

This world record benchmark result clearly demonstrates that the Sun SPARC Enterprise T5220 running the Solaris 10 05/08 and Java System Webserver 7.0 Update 3 can support thousands of concurrent secure and non secure web server sessions while allowing larger and more complex Java applications to be run and is an industry leader in web serving.

Cryptography performance is enhanced on the Sun SPARC Enterprise T5220 by using UltraSPARC T2's enhanced on-chip cryptographic hardware with Solaris 10's secured web service software feature. SPECweb2005's banking workload highlights the server's secure web server performance. The Sun SPARC Enterprise T5220 (70,000 SPECweb2005_Banking) performance was 35% greater than the 2 socket 3.16 GHz Quad-Core Xeon X5460 HP ProLiant DL380 G5 (51,840 SPECweb2005_Banking).

The HP Proliant DL580 G5 (40,046 SPECweb2005) used seventeen 1GbE networks versus Sun's three 10GbE and two 1GbE, providing much simplified administration.

At peak load, the Sun SPARC Enterprise T5220 had an average power consumption of 617 watts for the three SPECweb2005 benchmark workloads at steady-state.

The SWaP metric is a measure of server efficiency ratio that includes system performance, power and space consumption on a specific benchmark. (SWaP = Perf /[ Space (RU) x Watts ] )

Power-performance is computed as watt/performance. Since power-performance is related to price/performance they are both calculated with performance in the denominator.

Selected SPECweb2005 benchmark results as of 04/01/2008. Complete information at: http://www.spec.org website.

System Chips, Core
/Chip
CPU/GHz OS Web Server SPEC web
2005
Bank Ecom Supp
Sun SE T5220 1,8 US T2 1.4 Solaris 05/08 Sun JSWS 7.0 Update 3 41847 70000 58000 40000
HP PL DL580 G5 4,4 Xeon QC / 2.993 RedHat Linux Rock1.4.6/
JRock v1.3.1
40046 71104 55552 36032
Sun SE T5220 1,8 US T2/ 1.4 Solaris 05/08 Sun JSWS 7.0
Update 3
37001 63000 49500 36000
HP PL DL380 G5 2, 4 Xeon QC / 3.16 RedHat Linux Rock1.4.6/
JRock v1.3.1
29591 51840 46512 23816
HP PL DL580 G5 4, 4 Xeon QC/2.4 RedHat Linux Rock1.4.1/
JRock v1.2.0
26119 45056 37312 23488
Sun Fire T2000 1,8 US T1/1.4 Solaris 11/06 Sun JSWS 6.1
SP5 64b
16407 25812 24048 15768
Sun Fire T2000 1, 8 US T1/1.2 Solaris Sun JSWS 6.1
SP5 64b
14001 21500 21500 13160
Sun Fire T1000 1, 8 US T1/1.0 Solaris Sun JSWS 6.1
SP5 64b
10466 20000 16500 7700
IBM p5 550 2,2 POW5+/ 1.9 SuSE Linux Zeus4.3r1/
Tomcat5.5.9
7881 12240 11820 7500

Results in BLUE are the Sun SPARC Enterprise T5220 results currently under SPEC review.

Benchmark Description

SPECweb2005, is the latest industry standard benchmark for evaluating Web Server performance developed by SPEC. The benchmark simulates multiple user sessions accessing a Web Server and generating static and dynamic HTTP requests. The major features of SPECweb2005 are:

  • Measures simultaneous user sessions
  • Dynamic content: currently PHP and JSP implementations
  • Page images requested using 2 parallel HTTP connections
  • Multiple, standardized workloads: Banking (HTTPS), E-commerce (HTTP and HTTPS), and Support (HTTP)
  • Simulates browser caching effects
  • File accesses more accurately simulate today's disk access patterns

Example Disclosure Statement:

Sun SPARC Enterprise T5220 (8 cores, 1 chip) 41847 SPECweb2005, submitted to SPEC for review on April 8, 2008. HP ProLiant DL580 G5 (16 cores, 4 chips) 40046 SPECweb2005. HP ProLiant DL380 G5 (8 cores, 2 chips) 29591 SPECweb2005. HP ProLiant DL580 G5 (16 cores, 4 chips) 26119 SPECweb2005. Sun Fire T2000 (8 cores, 1 chip) 16407 SPECweb2005. Sun Fire T2000 (8 cores, 1 chip) 14001 SPECweb2005. Sun Fire T1000 (8 cores, 1 chip) 10466 SPECweb2005. IBM p5 550 (4 cores, 2 chips) 7881 SPECweb2005. SPEC, SPECweb reg tm of Standard Performance Evaluation Corporation. Results from www.spec.org as of Apr 7, 2008.

Sun SPARC Enterprise T5220 server power consumption taken from measurements made during the benchmark run. Power is average measured watts during benchmark run.

HP DL580 power consumption from HP Power Calculator system configured with 4 x2.93GHz processors, redundant PSU, 16 x 4GB DIMMs, 8 x 36GB SAS drives,1 x PCI card, 80% utilisation on 9/10/07: http://h30099.www3.hp.com/configurator/powercalcs.asp

HP DL380G5 power consumption from HP Power Calculator for system configured with 2 x X5460 3.16GHz processors, redundant PSU, 8 x 4GB DIMMs, 2 x HBAs and 2 x 146GB SAS drives, 80% utilisation on 01/09/07: http://h30099.www3.hp.com/configurator/powercalcs.asp

Results Summary

    Certified Results 41847 SPECweb2005
    Reference Date: April 7, 2008
    Systems: 1 x Sun SPARC Enterprise T5220
    Total Number Processors: 1 chip / 8 cores (8 threads/core)
    Processor/GHz of Server: Sun UltraSPARC T2 1.4 GHz
    Operating System: Solaris 10 05/08 + patches
    Software: Sun Java[TM] System Web Server 7.0 Update 3
    Sun Java[TM] Runtime Environment, Standard Edition 1.6.0_06-p

[1] Comments
Like this post? del.icio.us | furl | slashdot | technorati | digg
Comments:

You should just say "results," not "certified results." If you want to point out that results are peer reviewed by SPEC the easiest way to do that is simply to point to the result on spec.org.

Posted by Walter Bays on April 09, 2008 at 01:43 PM PDT #

Post a Comment:
Comments are closed for this entry.