Tuesday Apr 14, 2009
Today Sun announced world records for SPECfp2006: 50.4 on a 2-chip Nehalem (Intel Xeon X5570) Sun Blade X6270 as well as SPECint2006: 36.9 on a 2-chip Nehalem (Intel Xeon X5570) Sun Blade X6270.
Read more at: http://blogs.sun.com/jhenning/entry/sun_studio_trounces_intel_compiler.
Yes, even on servers based on the same CPUs as others, Sun can make a difference. Congrats to those on the Sun Studio Compiler team. They beats Intel's own compiler on this Intel chip by 20%, due to the optimization technologies found in the Sun Studio 12 Update 1 compiler.
See John's posting above for more info. On a different note, notice how much information Sun puts out our benchmarks - lots! Fun to look at IBM bloggers, some of whom spend 90% of their blog on "cute" and only 10% talking about benchmark results. Information is not ones enemy.
Disclosure Statement:
SPEC, SPECint, SPECfp reg tm of Standard Performance Evaluation Corporation. Results from www.spec.org as of 4/14/2009. Sun Blade X6270 (Intel Xeon X5570 / 2 chips / 8 cores) 50.4 SPECfp2006, IBM System p 570 (POWER6, 1 chip / 1 core) 24.9 SPECfp2006.
Tuesday Apr 14, 2009
This TPC-H result demonstrates that the Sun Fire X4275 server,
powered by 2 Quad-core 2.93 GHz Intel Nehalem X5570 processors, using
only 12 internal disks (SAS 300GB 15K RPM), achieved a QphH@1000GB of
23,365 with a price performance of $2.41. This is the best price
performance among all non-clustered server results at 1000GB.
Best price/performance among all TPC-H results at 1000GB, 70%
better than the previous best (Sun Fire X4500) and 75% better than
the previous second best ie. the HP DL585.
It is the Best 2-chip or 2-socket server result, even better than
many 4-sockets servers.
To put this result in perspective, the best non Sun single
server submission at 1000GB was the HP Superdome. The Superdome
achieved a QphH of 69,999 (about 3 times the Sun Fire X4275
performance) BUT: it required almost 100 times the number of
disks, more than 35 times the price and 8 times the number of cores
when compared to the Sun Fire X4275 configuration!
Once again, the Sun/SybaseIQ combination has produced a
self-contained (i.e. a server without any external storage or
external processing engines) data warehousing solution. Only Sun has
the hardware and expertise to produce such TPC-H results. To date,
Sun has published self-contained TPC-H results for each of the
100GB, 300GB, 1000GB and 3000GB scale-factors.
This is a extremely compact solution requiring only 2 rack units in total. Again contrast the Sun result
with the HP Superdome, using 97 storage arrays at 3 RU each plus
a 48 inch cabinet for the server.
Extremely efficient power consumption; peak power consumption
throughout the entire benchmark run was 825 Watts with an average
consumption of 750 Watts.
{humor: Any comments from HP or Dell or IBM why they never publish watts on any standard benchmarks with real size memory (i.e. anything above 16GB) ? } I'll take comments from incognito HP, IBM, or Dell employees below, as always. 
Performance Results
In order to put the Sun Fire X4275 TPC-H result in perspective,
the table below shows the top non-clustered TPC-H@1000 results from
Sun, Bull and HP in ascending order of $/QphH as of
April 14, 2009.
|
System
CPU
|
so/
co/
th
|
DB
|
QphH
|
$/QphH
|
Price
$USD
|
# Disks
|
Avail-
able
|
Data
Ratio
|
| Sun Fire X4275, 72GB
Intel X5540, 2.93GHz |
2/8/16 |
Sybase IQ |
23,365 |
2.41 |
56,263.91 |
12 |
4/14/09 |
3.5 |
| Sun Fire X4500, 64GB
AMD Opteron 2.8GHz |
2/4/4 |
Sybase IQ |
5,604 |
8.11 |
45,439 |
48 |
10/15/07 |
11.2 |
| HP DL585 G2, 32GB
AMD Opteron 2.8GHz |
4/8/8 |
SQL Server |
14,773 |
9.73 |
143,736 |
206 |
4/25/07 |
7.8 |
| Bull Novascale 3045, 64GB
Itanium 1.6GHz |
4/8/16 |
SQL Server |
12,087 |
12.56 |
151,870 |
160 |
3/6/07 |
5.7 |
| HP DL585 G1, 64GB
AMD Opteron 2.4GHz |
4/4/4 |
SQL Server |
10,493 |
13.83 |
145,264 |
164 |
3/2/06 |
6.4 |
| HP Superdome |
32/
64/
64 |
SQL Server |
69,999 |
28.69 |
2,008,168 |
1198 |
6/18/07 |
40.63 |
Legend:
so/co/th = sockets, cores, threads
QphH = Overall TPC-H Composite Metric (bigger is better).
$/QphH = TPC-H Price/Performance metric (smaller is better)
Data Ratio = Total disk to actual data ratio
Complete benchmark results may be found at http://www.tpc.org.
Benchmark Description
The results reported here were performed on a Sun Fire X4275
system and used Sybase IQ as the database manager. Sybase IQ is a
special product designed specifically for data warehousing
applications. Sybase IQ was developed as a totally separate product
from the more widely known Sybase database management system (Sybase
Adaptive Server).
The TPC-H benchmark is a performance benchmark established by the
Transaction Processing Council (TPC) to demonstrate Data
Warehousing/Decision Support Systems (DSS). TPC-H measurements are
produced for customers to evaluate the performance of various DSS
systems. These queries and updates are executed against a standard
database under controlled conditions. Performance projections and
comparisons between different TPC-H Database sizes (300GB, 300GB,
1000GB, 3000GB and 10000GB) are not allowed by the TPC.
TPC-H is a data warehousing-oriented, non-industry-specific
benchmark that consists of a large number of complex queries typical
of decision support applications. It also includes some insert and
delete activity that is intended to simulate loading and purging data
from a warehouse. TPC-H measures the combined performance of a
particular database manager on a specific computer system.
The main performance metric reported by TPC-H is called the TPC-H
Composite Query-per-Hour Performance Metric (QphH@SF, where SF is the
number of GB of raw data, referred to as the scale factor). QphH@SF
is intended to summarize the ability of the system to process queries
in both single and multi user modes. The benchmark requires reporting
of price/performance, which is the ratio of QphH to total HW/SW cost
plus 3 years maintenance. A secondary metric is the storage
efficiency, which is the ratio of total configured disk space in GB
to the scale factor.
The QphH composite metric is the Geometric Mean of 2 components:
(1) a single user component, called Power, and a (2) multi-user
component, called Throughput. Power is a performance
measurement of a single user stream of 22 queries, one batch insert
and one batch delete, all run serially. The Throughput metric,
instead, consists of essentially N concurrent Power streams (or
“users” submitting queries), where N is a minimum number
of required streams dependent upon the database size. For example, at
300GB, N must be at least 5 and at 300GB N must be at least 6. Both
Power and Throughput are calculated metrics and each is inversely
proportional to the queries elapsed time: thus the faster the queries
finish, the larger the metric becomes and the better the result.
Disclosure Statement:
TPC-H, QphH, $/QphH are registered trademarks of the Transaction
Processing Performance Council (TPC). More info at
http://www.tpc.org/. Sun Fire X4275
23,365@1000GB, $2.41/QphH@1000GB, available 4/14/09.
Results Summary
| Audited Results |
| |
Database Size: |
|
1000 GB (Scale Factor 1000) |
|
| |
TPC-H Composite: |
|
23,365.3 |
|
| |
Price/performance: |
|
$2.41 |
|
| |
Available |
|
4/14/09 |
|
| Number of Systems: |
|
1 |
|
| Total Number Processors: |
|
2 |
|
| Total Number of Cores |
|
8 |
|
| Total Number of Threads |
|
16 |
|
| Processor/MHz of Server: |
|
Intel Nehalem 2.93 GHz X5570 Quad Core |
|
| Storage: |
|
12 x 15K SAS drives (all internal) |
|
| Database: |
|
Sybase IQ 15 |
|
| Operating System: |
|
Solaris 10 |
|
| Total 3 year Cost: |
|
$56,263.91 |
|
| Other Performance Metrics |
|
|
|
| |
TPC-H Power: |
|
29,824.6 |
|
| |
TPC-H Throughput: |
|
18,304.9 |
|
| |
Database Load Time: |
|
5 Hr 39 Min |
|
| |
Storage Ratio: |
|
3.35 |
|
Tuesday Apr 14, 2009
The Sun Blade X6270 server demonstrates Sun's position of leadership
in Java based computing by publishing world record results
for the SPECjvm2008 benchmark. The Sun Blade X6270 server delivered a result of 317.13 SPECjvm2008 Base ops/m using the Sun Java
JDK 1.6.0_14 Performance Release with the OpenSolaris 2008.11
Operating System.
SPECjvm2008 Performance Chart (ordered by performance)
base: SPECjvm2008 Base ops/m (bigger is better)
peak: SPECjvm2008 Peak ops/m (bigger is better)
Ch/Co/Lc: Chips, Cores, Logical CPUs
| System |
Processors |
Performance |
| Ch |
Co |
Lc |
GHz |
Type |
base |
peak |
| Sun Blade X6270 |
2 |
8 |
16 |
2.93 |
X5570 QC |
317.13 |
- |
| Sun Fire X4450 |
4 |
24 |
24 |
2.66 |
X7450 6C |
283.79 |
- |
| Sun Fire X4450 |
4 |
16 |
16 |
2.93 |
X7350 QC |
260.08 |
- |
Benchmark Description
SPECjvm2008 (Java Virtual Machine Benchmark) is a benchmark suite for
measuring the performance of a Java Runtime Environment (JRE),
containing several real life applications and benchmarks focusing on
core java functionality. The suite focuses on the performance of the
JRE executing a single application; it reflects the performance of the
hardware processor and memory subsystem, but has low dependence on file
I/O and includes no network I/O across machines. The SPECjvm2008
workload mimics a variety of common general purpose application
computations. These characteristics reflect the intent that this
benchmark will be applicable to measuring basic Java performance on a
wide variety of both client and server systems.
SPEC also finds user experience of Java important, and the suite
therefore includes startup benchmarks and has a required run category
called base, which must be run without any tuning of the JVM to improve
the out of the box performance.
SPECjvm2008 benchmark highlights:
-
Leverages real life applications (like derby, sunflow, and javac) and
area-focused benchmarks (like xml, serialization, crypto, and
scimark).
-
Also measures the performance of the operating system and hardware in
the context of executing the JRE.
Disclosure Statement:
SPEC, SPECjvm reg tm of Standard Performance Evaluation Corporation.
Results as of 4/14/08 on http://www.spec.org.
Sun Blade X6270(2 chips, 8 cores) 317.13 SPECjvm2008 Base ops/m submitted
to SPEC for review.
Sun Fire X4450(4 chips, 24 cores) 283.79 SPECjvm2008 Base ops/m
Sun Fire X4450(4 chips, 16 cores) 260.08 SPECjvm2008 Base ops/m
System Configuration
|
Results
|
|
|
Performance:
|
|
317.13 SPECjvm2008 Base ops/m
|
|
|
Reference Date:
|
|
Apr 14, 2009
|
|
Systems:
|
|
Sun Blade X6270
|
|
Total Number Processors:
|
|
2
|
|
Processor/ GHz of Server:
|
|
Intel Xeon X5570 QC 2.93 GHz
|
|
Operating System:
|
|
OpenSolaris 2008.11
|
|
JVM:
|
|
Java HotSpot(TM) 64-Bit Server VM on Solaris, version 1.6.0_14 Performance Release
|
Monday Dec 08, 2008
The Sun Fire X4450 64GB with four 2.933 GHz Quad-core Xeon X7350
obtained 39,793 SPECweb2005 using only 883 watts. The Sun Fire X4450
was running SuSE Linux Enterprise Server 10 SP2 with Accoria Networks,
Rock Web
Server v1.4.7 and Rock JSP/Servlet Container v1.3.2. This result
demonstrates Sun Microsystems continued commitment
to deliver the high performing servers, regardless of operating system
environments.
The Sun Fire X4450 with 2.933GHz QC Xeons with 64GB of memory had an
average power consumption of 883 watts. The average
power consumption was measured for each of the three SPECweb2005
benchmark workloads during the steady state load period at peak load.
The power-performance for the Sun Fire X4450 is .022,
power-performance is computed as watt/performance. Since
power-performance is related to price/performance they are both
calculated with performance in the denominator.
This result
is competitive to the HP ProLiant DL580 G5, HP's 4 socket Xeon
X7350 offering.
The Sun Fire X4450 with 4x10Gb network links obtained similar
performance to the HP DL580 G5 with 16x1Gb network links. By
obtaining similar performance with fewer networks, IT management costs
and server power consumption are reduced.
Competitive Landscape
Selected SPECweb2005 benchmark results as of 12/04/2008. Complete
information at: http://www.spec.org
website.
System
OS |
Chips, Cores
/Chip |
CPU (GHz) |
Web Server |
SPEC web
2005 |
Bank |
Ecom |
Supp |
HP PL DL585 G5
RedHat Linux |
4, 4 |
AMD 8356 QC (2.3) |
Rock1.4.7/ JRock v1.3.2 |
43854 |
76032 |
62304 |
39456 |
Fujitsu RX600 S4
RedHat Linux |
4, 4 |
Xeon QC X7350 (2.93) |
Rock1.4.7/ JRock v1.3.2 |
42783 |
75008 |
59264 |
39040 |
Sun SE T5220
Sol 10 05/08 |
1, 8 |
US T2 (1.4) |
Sun JSWS 7.0 U3 |
41847 |
70000 |
58000 |
40000 |
HP PL DL580 G5
RedHat Linux |
4, 4 |
Xeon QC X7350 (2.93) |
Rock1.4.6/JRock v1.3.1 |
40046 |
71104 |
55552 |
36032 |
Sun Fire X4450
SLES 10 SP2 |
4, 4 |
Xeon QC X7350 (2.93) |
Rock1.4.7/ JRock v1.3.2 |
39793 |
70500 |
57000 |
34750 |
Sun Fire X4240
RedHat Linux |
2, 4 |
AMD QC 2360SE (2.5) |
Rock1.4.7/ JRock v1.3.2 |
32288 |
53048 |
50008 |
28120 |
HP PL DL580 G5
RedHat Linux |
4, 4 |
Xeon QC X7350 (2.93) |
Rock1.4.1/JRock v1.2.0 |
30261 |
52160 |
42048 |
28000 |
HP PL DL385 G5
RedHat Linux |
2, 4 |
AMD 2356 QC (2.3) |
Rock1.4.6/JRock v1.3.1 |
30007 |
50856 |
46020 |
25584 |
HP PL DL380 G5
RedHat Linux |
2,4 |
Xeon QC X5460 (3.16) |
Rock1.4.6/ JRock v1.3.1 |
29591 |
51840 |
46512 |
23816 |
IBM p5 550
SuSE Linux |
2, 2 |
DC Pow5+ (1.9) |
Zeus4.3r1/ Tomcat 5.5.9 |
7881 |
12240 |
11820 |
7500 |
Measured power consumption from Sun Fire X4450 64GB with 4 2.933GHz QC Xeon, smaller watt/performance is better.
| System |
Work- load Ave |
watt / perf |
Banking |
Ecomm |
Support |
Avg
Watts |
Min, Max
Watt |
Avg
Watt |
Min, Max
Watt |
Avg
Watt |
Min, Max
Watt |
| 1 x Sun Fire X4450 64GB 4-chip 2.93GHz QC Xeon |
883w |
.0222 |
903w |
781w, 923w |
881w |
620w, 906w |
867w |
734w, 897w |
| 1 x Sun Fire T5220 64GB 1-chip 1.4GHz US T2 |
617w |
.0147 |
|
|
|
|
|
|
| 1 x Sun Fire X4240 32GB 2-chip 2.5GHz QC Opteron |
521w |
.0161 |
|
|
|
|
|
|
At peak load, the Sun SPARC Enterprise T5220 with 64GB had an average power consumption of 617 watts for the three SPECweb2005 benchmark workloads at steady-state measured watts during benchmark run. A blog about this result at: http://blogs.sun.com/bmseer/entry/sun_s_even_faster_specweb2005
At peak load, the Sun SPARC Enterprise X4240 with 32GB had an average power consumption of 521 watts for the three SPECweb2005 benchmark workloads at steady-state measured watts during benchmark run.
You can look up competitive watts at:
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:
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:
http://h30099.www3.hp.com/configurator/powercalcs.asp
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 Fire X4450 (16 cores, 4 chips) 39793 SPECweb2005. HP ProLiant
DL585 G5 ( 16 core, 4 chips) 43854 SPECweb2005. Fujitsu Siemems
Primergy RX600 S4 (16 cores, 4 chips) 42783 SPECweb2005.
Sun SPARC Enterprise X5220 (8 cores, 1 chip) 41847 SPECweb2005.
HP ProLiant DL585 G5 (16 cores/4 chips) 40046 SPECweb2005 (re-run of
2.93GHz). Sun Fire X4240 (8 cores, 2 chips) 32288
SPECweb2005. HP ProLiant DL585 G5 (16 cores, 4 chips) 30261
SPECweb2005 (initial 2.93 GHz run). HP ProLiant DL385 G5 (8
cores, 2 chips) 30007
SPECweb2005. HP ProLiant DL380 G5 (8 cores, 2 chips) 29591
SPECweb2005. PRIMERGY TX300 S4 (8 cores, 2 chips) 28127
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 Dec 4, 2008.
Sun Fire X4450 server power consumption was taken from measurements
made during the benchmark runs of each workload at peak load, during
the steady state load period. The power is the average measured watts
during the benchmark run.
Results Summary
| Results |
39793 SPECweb2005 |
| Reference Date: |
Dec 4, 2008 |
| Systems: |
1 x Sun Fire X4450, 64GB |
| Total Number Processors: |
4 chips / 4 cores per chip |
| Processor/GHz of Server: |
4 x Xeon 7350 2.933 GHz |
| Operating System: |
SuSE Linux Enterprise Server 10 SP2
|
| Software: |
Rock Web Server v1.4.7 & Rock JSP/Servlet Container
v1.3.2
|
|
Java HotSpot[TM] 64-Bit Server VM on Linux, Version
1.6.0_06 Performance Release
|
Wednesday Nov 26, 2008
Over that the Zeus website, they posted information on Zeus Technology’s ZXTM (Zeus Extensible Traffic Manager).
Zeus shows performance data from their testing on QC Xeons, older DC Opterons, f5's Big-IP, Citrix, and Sun's CMT on HTTP throughput tests. What the data shows is the Sun Fire T5140 (1.2GHz, 2-socket) is the fastest 2-socket system.
Estimating from the graph the Sun Fire T5140 looks to have almost 3x better price peformance than f5's BIG-IP VIPRION and 2.5x better Citrix.
You can read more about it at:
http://www.zeus.com/news/press_articles/zeus-price-performance-press-release.html?gclid=CLn4jLuuk5cCFQsQagod7gTkJA
As they say, "Voice-over-IP (VoIP) services are converging on SIP and RTP for signaling and real-time media delivery respectively."
I'll post more details on this test as I find them.
Onwards to that most 'merican of holidays, T-day. Enjoy your holiday!
Wednesday Jun 18, 2008
Today Sun announced its powerful new Sun Blade X6450 server module at the International Supercomputing Conference (ISC) in Dresden, Germany (blog pics here).
sun.com Sun Blade X6450 features
Sun writes: "The Sun Blade X6450 is the newest in the 6000 series server modules, and brings the Sun Constellation System to the next level of performance through enhanced features that include up to four Intel Xeon dual- or quad-core processors, an optional 16GB Compact Flash storage subsystem, 24 DIMM slots and 110Gbps I/O throughput. This potent technology combination can deliver up to seven teraflops of performance per fully populated Sun Blade 6048 chassis and up to 71% more compute cores."
I'm sure http://blogs.sun.com/HPC/ will write more about ISC in Dresden.
Tuesday Feb 26, 2008
{update} There is a lot of information about MySQL and Sun at http://www.sun.com/mysql
In addition, I've put together a list of several blogs on MySQL performance.
* a very interesting results that compares Solaris Open-source stack (OS, DB, Web, Virtualizaion) on a 1-chip UltraSPARC T2 server and beating a proprietary stack on a 4-chip QC Xeon. Also measured actual watts and costs. Seems real configurations of HP DL580's draw lots of watts:
http://blogs.sun.com/ritu/entry/mysql_benchmark_us_t2_beats
* an ERP result using MySQL with SugarCRM:
http://blogs.sun.com/vanga/entry/scaling_sugarcrm_with_mysql_on
* great information about tuning MySQL on linux and some performance results:
http://blogs.sun.com/allanp/entry/tuning_mysql_on_linux
* nice writeup on InnoDB on SysBench:
http://blogs.sun.com/realneel/entry/tuning_mysql_innodb_for_sysbench
For a For a variety of things on MySQL see:
http://blogs.sun.com/barton808/entry/mysql_done_deal_talking_with
Tuesday Feb 26, 2008
Getting ready to head off for lunch and I took off my blinders and I see
all of the MySQL announcements. There are even several blogs on MySQL performance. Already some very interesting things coming from bringing MySQL into Sun.
* a very interesting results that compares Solaris Open-source stack (OS, DB, Web, Virtualizaion) on a 1-chip UltraSPARC T2 server and beating a proprietary stack on a 4-chip QC Xeon. Also measured actual watts and costs. Seems real configurations of HP DL580's draw lots of watts:
http://blogs.sun.com/ritu/entry/mysql_benchmark_us_t2_beats
* an ERP result using MySQL with SugarCRM:
http://blogs.sun.com/vanga/entry/scaling_sugarcrm_with_mysql_on
* great information about tuning MySQL on linux and some performance results:
http://blogs.sun.com/allanp/entry/tuning_mysql_on_linux
For a For a variety of things on MySQL see:
http://blogs.sun.com/barton808/entry/mysql_done_deal_talking_with
Monday Feb 18, 2008
The Sun Fire X4150 server equipped with 2 Quad-core Intel Xeon processors
obtained World Record x86 single-JVM and x86 2 chip multi-JVM
results on the SPECjbb2005 benchmark. Enhancements to the JVM had a major impact on performance.
The Sun Fire X4150 with 2 Intel X5460 quad-core processors and running Sun
J2SE 1.6.0_05-p achieved x86 2-chip World Record performance
of 303297 SPECjbb2005 bops,
75824 SPECjbb2005 bops/JVM on the SPECjbb2005 benchmark for multi-JVM results.
The Sun Fire X4150 with 2 Intel X5460 quad-core processors and running Sun
J2SE 1.6.0_05-p achieved x86 World Record performance
of 277585 SPECjbb2005 bops,
277585 SPECjbb2005 bops/JVM for a single JVM
on the SPECjbb2005 benchmark.
Using the same processor, the Sun Fire X4150 with Solaris 10 and
Java HotSpot(TM) 32-Bit Server, beat the results of Dell and Lenovo
which used Windows and BEA JRockit on the multi-JVM test.
The Sun Fire X4150 running the single-JVM SPECjbb2005 test easily
beat all x86 results, topping the Dell R200 by 1.98X, the Fujitsu BX620
by 2.0X and the SGI XE240 by nearly 2.1X.
SPECjbb2005 Performance Chart (ordered by performance, bops: SPECjbb2005 Business Operations per Second (bigger is better), selected results... my best guess at the top ones for the engineers who act like lawyers you can go to www.spec.org to see all results as it clearly states at the bottom.
| System |
Processors |
Performance |
| Chip |
Core |
Thr |
GHz |
Type |
SPEC-
jbb-
2005
bops |
JVMs |
SPEC-
jbb-
2005
bops/JVM |
| Multi-JVM, 2-Chip x86 Results (selected top see note above) |
| Sun Fire X4150 |
2 |
8 |
8 |
3.16 |
X5460 |
303297 |
4 |
75824 |
| Dell PowerEdge 2950 |
2 |
8 |
8 |
3.16 |
X5460 |
303130 |
4 |
75783 |
| Lenovo R515 |
2 |
8 |
8 |
3.16 |
X5460 |
294716 |
4 |
73678 |
| Single-JVM x86 Results (selected top, see note above) |
| Sun Fire X4150 |
2 |
8 |
8 |
3.16 |
X5460 |
277585 |
1 |
277585 |
| Dell PowerEdge R200 |
1 |
4 |
4 |
2.66 |
X3230 |
140220 |
1 |
140220 |
| Fujitsu BX620 |
2 |
4 |
4 |
3.0 |
5160 |
138388 |
1 |
138388 |
| SGI Altix XE240 |
2 |
4 |
4 |
3 |
5160 |
134561 |
1 |
134561 |
| Dell PowerEdge 2950 |
2 |
4 |
4 |
3 |
5160 |
130589 |
1 |
130589 |
Complete benchmark results may be found at the SPEC benchmark website http://www.spec.org.
Benchmark Description
SPECjbb2005 (Java Business Benchmark) measures the performance of a Java implemented application tier (server-side Java). The benchmark is based on the order processing in a wholesale supplier application. The performance of the user tier and the
database tier are not measured in this test. The metrics given are number of SPECjbb2005 bops (Business Operations per Second) and SPECjbb2005 bops/JVM (bops per JVM instance).
Disclosure Statement:
SPEC, SPECjbb reg tm of Standard Performance Evaluation
Corporation. Results as of 2/8/2008 on www.spec.org.
Sun Fire X4150 (2 chips, 8 cores) 303297 SPECjbb2005 bops,
75824 SPECjbb2005 bops/JVM;
Dell PowerEdge 2950 (2 chips, 8 cores) 303130 SPECjbb2005 bops,
75783 SPECjbb2005 bops/JVM;
Lenovo R515 (2 chips, 8 cores) 294716 SPECjbb2005 bops,
73678 SPECjbb2005 bops/JVM.
SPEC, SPECjbb reg tm of Standard Performance Evaluation
Corporation. Results as of 2/8/2008 on www.spec.org.
Sun Fire X4150 (2 chips, 8 cores) 277585 SPECjbb2005 bops,
277585 SPECjbb2005 bops/JVM;
Dell PowerEdge R200 (1 chip, 4 cores) 140220 SPECjbb2005 bops,
140220 SPECjbb2005 bops/JVM;
Fujitsu BX620 (2 chips, 4 cores) 138388 SPECjbb2005 bops,
138388 SPECjbb2005 bops/JVM;
SGI Altix XE240 (2 chips, 4 cores) 134561 SPECjbb2005 bops,
134561 SPECjbb2005 bops/JVM;
Dell PowerEdge 2950 (2 chips, 4 cores) 130589 SPECjbb2005 bops,
130589 SPECjbb2005 bops/JVM.
Results Summary
| Reference Date: |
|
Feb 8, 2008 |
| Multi-JVM |
|
303297 SPECjbb2005 bops, 75824 SPECjbb2005 bops/JVM |
| Single-JVM |
|
277585 SPECjbb2005 bops, 277585 SPECjbb2005 bops/JVM |
| System: |
|
Sun Fire X4150 |
| Processor: |
|
2 x Intel X5460 3.166 GHz |
| Operating System: |
|
Solaris 10 8/07 |
| JVM: |
|
Java HotSpot(TM) 32-Bit Server, Version 1.6.0_05-p |
Monday Jan 07, 2008
You may have missed this writeup about UltraSparc T2 and Tigerton Tests which looked at low-level memory access measurements: http://blogs.sun.com/psa/entry/ultrasparc_t2_sun
A quote from a Sun employee I like... "You can only compute as fast as you can move data"
Friday Oct 26, 2007
Sun gets two new World Records for scientific desktop performance
on the Sun Ultra 24 (single 3.0 GHz Intel DC Xeon E6850)
and the 64-bit SuSE Linux SLED 10 operating system.
Results obtained from the most current competitive platforms have
been recently posted for two different Mathematica 6 benchmarks:
- The Wolfram (Mathematica ISV) Benchmark
- The Independent (Mathematica MMA6.0.nb ) Benchmark
Although both of the Mathematica 6 benchmark test suites
contains 15 test cases these test cases are different
and the two test suites are separate and distinct from each other.
The Ultra 24 beats all results currently listed at both benchmark sites.
The Wolfram (Mathematica ISV) benchmark the Ultra 24
beats other current Intel Xeon (Woodcrest) dual core platforms
(3.0 GHz & 2.66 GHz), Intel based Apple MAC desktops. Itanium 2 platforms,
Pentium 4 platforms, and the IBM Power based platforms.
Alternatively, the independent Mathematica MMA6.0 notebook
benchmark the Ultra 24 beats posted results from primarily
current competitive Apple MAC desktops:
MacPro, MacBook, iMac, and Apple Powerbook G4
Results for both benchmark test suites are shown in the Two Tables
below under "Competitive Landscape"
Table 1. The Wolfram (Mathematica ISV) Benchmark
Summary results as in the installed Mathematica 6 Data Base.
This is the latest version of Mathematica timing tests.
Overall performance in 15 test calculations (Bigger is better)
The current reference is a machine with a 2.4 GHz Pentium 4 processor
| PLATFORM | Score
|
| 1 socket DC 3GHz Intel Xeon DC E6850 SLED 10 SP 1 Ultra 24 |
3.266 |
| 2-socket DC 3GHz Intel Xeon 5160 MS 32-bit |
2.84 |
| 2-socket DC 3.2 GHz Opteron 2224 Ultra 40 M2 64-bit SLED 10 32 GB |
2.736 |
| 2-socket DC 3.2GHz Opteron 2224 Ultra 40 M2 32-bit Windows XP SP2 8GB |
2.45 |
| 2 socket DC 2.66 GHz Intel Xeon 64-bit Apple MAC 10.4.8 |
2.14 |
| 2 socket QC 1.6 GHz Intel Xeon 5310 32-bit Cent OS Linux |
1.88 |
| 2 socket DC 2.5 GHz G5 Apple MAC OS 10.4.8 32-bit |
1.22 |
| 1 socket 2.4 GHz Pent. 4 MS Win XP 32-bit |
1.00 |
The Independent (Mathematica MMA6.0.nb ) Benchmark
Summary results as listed at the independent Mathematica MMA6
http://smc.vnet.net/timings60.html website.
This is the latest version of the "Mathematica MMA" timing tests.
Overall performance in 15 test calculations (Bigger is better)
The current reference machine is one with a 2.33GHz Intel Core 2 Duo processor.
| PLATFORM | Score
|
| Sun Ultra 24 3.0 GHz DC Intel E6850 8GB SuSE 10 SP 1 |
1.27505 |
| MacPro, 3.0GHz Intel Core2 Duo, 4GB, MacOS 10.4.9 [4] |
1.25404 |
| AMD Athlon 64 FX-74, 3.0GHz Socket F (1207 FX) DSDC, Windows [5] |
1.14464 |
| iMac, 2.33GHz Intel Core2 Duo, 3GB, MacOS 10.4.9 [2] |
1.00338 |
| MacBook Pro, 2.33GHz Intel Core2 Duo, 2 GB RAM, MacOS X 10.4.9 [1] |
1.00105 |
| MacBook, 2GHz Intel Core2 Duo, 2GB, MacOS 10.4.10 [6] |
0.880472 |
Benchmark Description
The Wolfram (Mathematica ISV) Benchmark
The Wolfram (Mathematica ISV) benchmark is a revised one
that now comes imbedded in the latest release of Mathematica (currently V6.0)
along with a database of results from current hardware vendor platforms.
This benchmark was developed by Schoeller Porter, one of the principlal
developers of Mathematica. He described the benchmark as follows:
This is the standard benchmark suite for Mathematica, initially introduced
in Mathematica 5.1 (as MathematicaMark2004). It includes both workstation
and parallel benchmarks. The parallel benchmark is automatically invoked
when the Parallel Computing Toolkit is loaded
and compute kernels are available.
It is actively developed, and MathematicaMark 6.0 is the current version.
The 15 Task benchmark includes:
Benchmark Name: MathematicaMark6
Full Version Number:6.0.1
Date: September 14, 2007
Benchmark Result: 3.266
Total Time 26.39
Results:
Data Fitting: 1.273
Digits of Pi: 0.488
Discrete Fourier Transform: 0.765
Eigenvalues of a Matrix: 2.059
Elementary Functions: 3.645
Gamma Function: 0.368
Large Integer Multiplication: 0.734
Matrix Arithmetic: 2.798
Matrix Multiplication: 3.062
Matrix Transpose: 1.298
Numerical Integration: 2.017
Polynomial Expansion: 1.352
Random Number Sort: 1.506
Singular Value Decomposition: 2.346
Solving a Linear System: 2.679
Output
Cell Change Times->{3.398799503863311*^9
The Independent (Mathematica MMA6.0.nb ) Benchmark
The Mathematica MMA 6 benchmark is a widely recognized benchmark.
The tasks are representative important scientific
computing desktop activities.
This benchmark was developed by karl.unterkofler@fh-vorarlberg.ac.at
The benchmark consists of 15 tasks.
Disclosure Statement:
Mathematica MMA 6 Scientific Benchmark Sun Fire Ultra 24 score: 1.27505. Mathematica is a reg
tm of Wolfram Research, Inc. results as of 10/23/07 on http://smc.vnet.net/timings60.htmlResults Summary
The Sun Ultra 24 workstation gives the best
desktop scientific computing performance as demonstrated with both
the The Wolfram (Mathematica ISV) Benchmark and the
The Independent (Mathematica MMA6.0.nb ) Benchmark.
Both of these 15 task benchmarks consists of operations that are representative
of computing a variety of scientific funtions.
| Reference Date | 23 October 2007 |
| |
| The Wolfram (Mathematica ISV) Benchmark |
| Platform | Sun Ultra 24 Workstation |
| Total Number Processors | 1 |
| Processor/GHz of Workstation | Intel DC E6850/3.0 GHz |
| Memory | 4x2 GB DDR2 667 MHz dimms |
| Operating System | 64-bit SUSE SLED 10 SP 1 |
| Graphics | nVidia Quadro FX 1700 framebuffer |
| Disks | 2x146 GB 15K rpm SAS striped |
| Software |
Mathematica 6 (Scientific Application)
Wolfram (ISV) Benchmark |
| Composite Score | 3.266 |
| |
| The Independent (Mathematica MMA6.0.nb ) Benchmark |
| Platform | Sun Ultra 24 Workstation |
| Total Number Processors | 1 |
| Processor/GHz of Workstation | Intel DC E6850/3.0 GHz |
| Memory | 4x2 GB DDR2 667 MHz dimms |
| Operating System | 64-bit SUSE SLED 10 SP 1 |
| Graphics | nVidia Quadro FX 1700 framebuffer |
| Disks | 2x146 GB 15K rpm SAS striped |
| Software |
Mathematica 6 (Scientific Application)
The Independent (Mathematica MMA6.0.nb ) Benchmark |
| Composite Score | 1.27505 |
Wednesday Oct 24, 2007
The Sun Ultra 24 desktop sets a world record in the MCAD market.
The Ultra 24 beats competitive platforms from Dell, IBM, and HP.
The single socket Ultra 24 can use either Intel dual-core and quad-core
processors. The Sun Ultra 24 demonstrates both excellent performance
and $/performance.
Pro/E is leading software MCAD system. Most major MCAE ISV applications have
integration with Pro/E. Pro/E is used in a variety of different disciplines
such as automotive, aircraft, aerospace, marine, oil&gas, earth moving,
biomedical, heavy industry, atomic energy, etc.
Sun supports Pro/E on Opteron-based desktop platforms and Xeon-based platforms.
Pro/E users appreciate Solaris for its maturity, reliability, suberb
maintenanace and comprehensive well developed network features. This is
a benefit for many engineering corporations that have distributed design.
The OCUS V5 benchmark has a 32-bit "Normal" benchmark
and a newer 64-bit "Large Memory" benchmark to show performance on larger new workloads.
The 32-bit "Normal" OCUS V5 benchmark and World Record Ultra 24 Performance
The Sun Ultra 24 (3GHz QC Intel QX6850 Xeon processor, 8GB memory,
an nVidia Quadr0 FX 5600 framebuffer, 2x15K SAS striped drives under 64-bit
Win 2003 SP 2 XP 64-Ed. sets a new MCAD world record running
the Pro/E Wildfire OCUS V5 32-bit "Normal" benchmark
beating all "legitimate" hardware vendors with results currently posted
at the OCUS V5 www.proesite.com benchmark website.
Reruns on the same Ultra 24 platform but with a 3GHz DC Intel
Xeon E6850 processor also with nVidia Quadro FX 5600 produced essentially
identical world record results as obtained in the initial runs with a
with a 3GHz QC E6850 processor.
Further reruns on the same Ultra 24 platform with a 3GHz DC Xeon
E6850 but with an nVidia Quadro FX 1700 produced essentially
identical world record results as obtained in the initial runs with a
3GHz QC QX6850 processor.
These results obtained with Pro/E WF 3 are better than any others posted
at the Pro/E Wildfire OCUS V5 "Normal" benchmark website by "legitimate"
harware vendors.
The top most competition comes from current Dell and HP desktop
platforms both with the recent dual-core 3GHz Woodcrest 5160
Intel processors or the Intel Core2 Duo Extreme processors
The 64-bit "Large Memory" OCUS V5 benchmark and World Record Ultra 24 Performance
-
The Sun Ultra 24 with a 3GHz DC Intel Xeon E6850, 8GB memory,
an nVidia Quadr0 FX 1700 framebuffer, 2x15K SAS striped drives under 64-bit
Win 2003 SP 2 XP 64-Ed. sets a new MCAD world record running
the Pro/E Wildfire OCUS V5 64-bit "Large Memory" benchmark
-
Reruns on the same Ultra 24 again with a 3GHz DC Intel Xeon E6850
but with the NVidia Quadro FX 5600 framebuffer instead
of the NVidia Quadro FX 1700 also produced essentailly the same
world record results.
-
Further reruns on the same Ultra 24 platform but now with a 3GHz QC Intel
QX6850 processor (same nVidia Quadro FX 5600 framebuffer)
produced essentially identical world record results as obtained
in the initial runs with a 3GHz DC E6850 Xeon and an nVidia Quadro
FX 1700 framebuffer.
-
These results obtained with Pro/E WF 3 are better than any others posted
at the Pro/E Wildfire OCUS benchmark website.
The top most competition comes from current Dell and HP desktop
platforms both equipped with the recent dual core 3GHz Woodcrest 5160
Intel processors or the Intel Core2 Duo Extreme processors
PRO/E WILDFIRE MCAD OCUS V5 32-bit "NORMAL" BENCHMARK Selected results are
run times in seconds, smaller is better
Ultra 24 vs. Topmost Current Posted Competitive Result
|
|
Time (in seconds) |
|
| Platform |
Processor |
Total |
Graphics |
CPU |
Disk I/O |
OS |
| Ultra 24 |
1x3.0GHz QC Intel QX6850 |
1228 |
664 |
563 |
91 |
Win 64 XP |
| Dell Prec 390 |
1x2.93GHz Intel Core2 X6800 |
1285 |
692 |
591 |
95 |
Win 64 XP |
PRO/E WILDFIRE MCAD OCUS V5 64-bit "Large Memory" BENCHMARK
Selected results are
run times in seconds, smaller is better
Ultra 24 vs. Topmost Current Posted Competitive Result
|
|
Time (in seconds) |
|
| Platform |
Processor |
Total |
Graphics |
CPU |
Disk I/O |
OS |
| Ultra 24 |
1x3.0GHz DC Intel E6850 |
2809 |
877 |
1926 |
352 |
Win 64 XP |
| Dell Prec. 490 |
1x3.0GHz DC Intel 5160 |
3026 |
1094 |
1925 |
341 |
Win 64 XP |
For results see OCUS website: http://www.proesite.com
Results Summary
PRO/E WILDFIRE MCAD OCUS V5 32-bit "NORMAL" BENCHMARK
|
|
| Submitted Results | 32-bit "Normal" OCUS V5 Benchmark |
| Reference Date | 23 October 2007 |
| Platform | Sun Ultra 24 Workstation |
| Total Number Processors | 1 |
| Processor/GHz of Workstation | Intel QC QX6850/3.0 GHz |
| Memory | 4x2 GB DDR2 667MHz dimms |
| Operating System | Win 2003 SP 2 64 Ed. |
| Graphics | nVidia Quadro FX 5600 framebuffer |
| Disks | 2x146 GB 15K rpm SAS striped |
| Software |
Pro/E Wildfire 3 (MCAD Application) |
| OCUS V5 32-bit "Normal" Benchmark |
| Total Elapsed Time | 1228 seconds |
| Total CPU Time | 563 seconds |
| Total Graphics Time | 664 seconds |
| Total Disk I/O Time | 91 seconds |
|
PRO/E WILDFIRE MCAD OCUS V5 64-bit "Large Memory" BENCHMARK
|
|
| Submitted Results | 64-bit OCUS V5 Benchmark |
| Reference Date | 23 October 2007 |
| Platform | Sun Ultra 24 Workstation |
| Total Number Processors | 1 |
| Processor/GHz of Workstation | Intel DC E6850/3.0 GHz |
| Memory | 4x2 GB DDR2 667MHz dimms |
| Operating System | Win 2003 SP 2 64 Ed. |
| Graphics | nVidia Quadro FX 1700 framebuffer |
| Disks | 2x146 GB 15K rpm SAS striped |
| Software |
Pro/E Wildfire 3 (MCAD Application) |
| OCUS V5 64-bit "Large Memory" Benchmark |
| Total Elapsed Time | 2809 seconds |
| Total CPU Time | 1926 seconds |
| Total Graphics Time | 877 seconds |
| Total Disk I/O Time | 352 seconds |
|
Thursday Aug 09, 2007
Postscript:
Be careful when comparing performance results, as an example look at
a comment in yesterday's
"Can I use 64 threads in a chip?" posting. At
least this comment pointed out that you can use 4-8 threads in 2 chip Intel-based systems, but it was really trying to
be a stab at UltraSPARC Performance. Here was the comment:
One really needs to look at the complete data on those .pdf's
to make a fair comparison (also in the disclosure statement
below).
First: The T2000 SAP-SD used a 1.2GHz UltraSPARC T1, Sun now ships faster 1.4GHz UltraSPARC T1, and has announced 1.4GHz UltraSPARC T2. The 1.4GHz T2 has double the threads of that 1.4GHz (double the computational power).
Second: The T2000 SAP-SD result was submitted in Dec 2005, at that time it
was near the performance of the expensive 4-way POWER5 IBM p550.
Third: The 2-chip Dual-core Xeon SAP-SD result above was
submitted 18 months after the T2000 SAP-SD result.
Fourth: Different versions of the benchmark. The 2-chip
Dual-core Xeon was run with ECC 6.0 (not SAP 5.0). The a newer version
of the benchmark takes more computational work to produce the same results.
Dual-core SAP-SD result was also run with Solaris 10 on Xeon, how cool is that!
Fifth: The 2-chip quad-core Xeon SAP-SD result above was
submitted 19 months after the T2000 SAP-SD result.
Sixth: The Sun result used open-source MySQL MaxDB database,
how cool is that! The Xeon results used Oracle or MicroSoft SQL Server.
postscript:
Sun latter used Oracle, others suggested US T1 has some sort of silly database limitation - NOT TRUE!
You'll see more results soon.
Triffids, as a reminder if you work for a partner company of SAP you must
put the following disclosures when you post results. If you are not
they you don't need to put this in, but as you can see the data in
it would have allowed you to make a better comparison of systems.
Don't worry I'm not asking you to identify yourself at all.
Disclosure Statement:
Two-tier SAP ECC 5.0 Standard Sales and Distribution (SD) benchmark Sun Fire T2000 (1-way, 1 proc, 8 cores, 32 threads) 1x 1.2 GHz UltraSPARC T1, 32 GB mem, 950 SD benchmark users, 1.91 sec avg response time, Cert#2005047., MaxDB 7.5 database, Solaris 10; Two-tier SAP ECC 5.0 Standard Sales and Distribution (SD) benchmark IBM System eServer p5 550 (4-way, 4 procs, 4 cores, 8 threads) 4x 1.9 GHz POWER5+, 32GB mem, 1,000 SD benchmark users, 1.97s avg resp time, Cert#2005040, IBM DB2 Universal Database 8.2.2, SuSE Linux Enterprise Server 9;
Two-tier SAP ECC 6.0 Standard Sales and Distribution (SD) benchmark Fujitsu Siemens Computers PRIMERGY Model BFi20 S2 (2 procs, 4 cores, 4 threads) 2x Intel Xeon 5160, 3.0 GHz, 16GB mem, 1,020 SD benchmark users, 1.94s avg resp time, Cert#2007031, Oracle 10g, Solaris 10;
Two-tier SAP ECC 6.0 Standard Sales and Distribution (SD) benchmark Fujitsu Siemens Computers PRIMERGY Model TX300 S3 (2 procs, 8 cores, 8 threads) 4x Quad-Core Intel Xeon Processor X5355 2.66 GHz, 32GB mem, 1865 SD benchmark users, 1.99s avg resp time, Cert#2007025, SQL Server 2005, Windows Server 2003 Enterprise Edition; SAP, R/3, mySAP reg TM of SAP AG in Germany and other countries. More info www.sap.com/benchmark.
I edited in:
2 processors into Quad-Core Intel Xeon Processor X5355 2.66 GHz
...and..
32 threads to the Sun Fire T2000, 1 processor / 8 cores
...in order to make the comparisons more consistent.
|
|
|
|
I suppose that even with Sun publishing their firs...