SPECjbb2015: SPARC T8-1 World Record Single Chip Multi-JVM Result

September 19, 2017 | 5 minute read
Brian Whitney
Principal Software Engineer
Text Size 100%:

Oracle's SPARC T8-1 server, using Oracle Solaris and Oracle JDK, produced a world record single-chip SPECjbb2015-MultiJVM benchmark result. This benchmark was designed by the industry to showcase Java performance in the Enterprise. Performance is expressed in terms of two metrics, max-jOPS which is the maximum throughput number and critical-jOPS which is critical throughput under service level agreements (SLAs).

  • The SPARC T8-1 server achieved 153,532 SPECjbb2015-MultiJVM max-jOPS and 89,980 SPECjbb2015-MultiJVM critical-jOPS on the SPECjbb2015 benchmark. On the SPARC T8-1 server's 32 cores, these rates are 4,798 SPECjbb2015-MultiJVM max-jOPS per core and 2,812 SPECjbb2015-MultiJVM critical-jOPS per core.

  • The SPARC T8-1 server delivered 1.5 times more SPECjbb2015-MultiJVM critical-jOPS performance per core than the SPARC T7-1 server which uses SPARC M7 processors. The SPARC T8-1 server also produced 1.3 times more SPECjbb2015-MultiJVM max-jOPS performance per core compared to the SPARC T7-1 server.

  • The SPARC T8-1 server delivered 2.7 times more SPECjbb2015-MultiJVM critical-jOPS performance per core than the Cisco UCS C240 M5 server using Intel Skylake processors (Intel Xeon Platinum 8180 Processor). The SPARC T8-1 server also produced 1.5 times more SPECjbb2015-MultiJVM max-jOPS performance per core compared to the same Cisco UCS C240 M5 server.

From SPEC's press release: "The SPECjbb2015 benchmark is based on the usage model of a worldwide supermarket company with an IT infrastructure that handles a mix of point-of-sale requests, online purchases, and data-mining operations. It exercises Java 7 and higher features, using the latest data formats (XML), communication using compression, and secure messaging."

Performance Landscape

Single chip results of SPECjbb2015 MultiJVM from www.spec.org as of August 28, 2017 and this report (for SPARC T8-1 results). Results ordered by by SPECjbb2015-MultiJVM max-jOPS, bigger is better.

SPECjbb2015
MultiJVM Results
System Performance Perf/Core Environment
max crit max crit
SPARC T8-1
1 x SPARC M8
(5.06 GHz, 1x 32core)
153,532 89,980 4,798 2,812 Oracle Solaris 11.3
JDK 8u144
SPARC T7-1
1 x SPARC M7
(4.13 GHz, 32core)
120,603 60,280 3,769 1,884 Oracle Solaris 11.3
JDK 8u66
HP ProLiant DL380 Gen10
1 x Intel Xeon Platinum 8180
(2.50 GHz, 1x 28core)
84,142 25,431 3,005 908 SUSE Linux Enterprise Server 12 SP2
JDK 8u131
Lenovo ThinkSystem SR650
1 x Intel Xeon Platinum 8180
(2.50 GHz, 1x 28core)
83,909 26,145 2,997 934 SUSE Linux Enterprise Server
JDK 8u131
HP ProLiant DL380 Gen10
1 x Intel Xeon Platinum 8180
(2.50 GHz, 1x 28core)
74,086 62,592 2,646 2,235 Red Hat Enterprise Linux Server 7.3
JDK 8u131
 

Best SPECjbb2015-MultiJVM max-jOPS per core results (all chip counts) from www.spec.org as of August 28, 2017 and this report (for SPARC T8-1 results). Results ordered by Perf/Core, bigger is better.

SPECjbb2015
MultiJVM Results
System Performance Perf/Core Environment
max crit max crit
SPARC T8-1
1 x SPARC M8
(5.06 GHz, 1x 32core)
153,532 89,980 4,798 2,812 Oracle Solaris 11.3
JDK 8u144
Fujitsu SPARC M12-2S
1 x SPARC64 XII
(4.25 GHz, 1x 12core)
54,434 34,771 4,536 2,898 Oracle Solaris 11.3
JDK 8u121
IBM Power S812LC
1 x POWER8
(2.92 GHz, 10core)
44,883 13,032 4,488 1,303 Ubuntu 14.04.3
J9 VM
SPARC S7-2
2 x SPARC S7
(4.26 GHz, 2x 8core)
65,790 35,812 4,112 2,238 Oracle Solaris 11.3
JDK 8u92
SPARC T7-1
1 x SPARC M7
(4.13 GHz, 32core)
120,603 60,280 3,769 1,884 Oracle Solaris 11.3
JDK 8u66
Cisco UCS C240 M5
2 x Intel Xeon Platinum 8180
(2.50 GHz, 2x 28core)
179,534 58,094 3,206 1,037 SUSE Linux Enterprise Server 12 SP2
JDK 8u131
Lenovo ThinkSystem SR650
2 x Intel Xeon Platinum 8180
(2.50 GHz, 2x 28core)
177,561 54,418 3,170 972 SUSE Linux Enterprise Server
JDK 8u131
Huawei FusionServer 2288H V5
2 x Intel Xeon Platinum 8180
(2.50 GHz, 2x 28core)
175,588 48,977 3,135 874 Red Hat Enterprise Linux
JDK 8u131
HP ProLiant DL380 Gen10
1 x Intel Xeon Platinum 8180
(2.50 GHz, 1x 28core)
84,142 25,431 3,005 908 SUSE Linux Enterprise Server 12 SP2
JDK 8u131
Lenovo ThinkSystem SR650
1 x Intel Xeon Platinum 8180
(2.50 GHz, 1x 28core)
83,909 26,145 2,997 934 SUSE Linux Enterprise Server
JDK 8u131
Huawei RH2288H V3
2 x Intel Xeon E5-2699 v4
(2.2 GHz, 2x 22core)
121,381 38,595 2,759 877 Red Hat 6.7
JDK 8u92
HP ProLiant DL360 Gen9
2 x Intel Xeon E5-2699 v4
(2.2 GHz, 2x 22core)
120,674 29,013 2,743 659 Red Hat 7.2
JDK 8u74
Huawei RH2288H V3
2 x Intel Xeon E5-2699 v3
(2.3 GHz, 2x 18core)
98,673 28,824 2,741 801 Red Hat 6.7
JDK 8u92
HP ProLiant DL380 Gen10
1 x Intel Xeon Platinum 8180
(2.50 GHz, 1x 28core)
74,086 62,592 2,646 2,235 Red Hat Enterprise Linux Server 7.3
JDK 8u131
SPARC T5-2
2 x SPARC T5
(3.6 GHz, 2x 16core)
80,889 37,422 2,528 1,169 Oracle Solaris 11.2
JDK 8u66
HP ProLiant DL380 Gen9
2 x Intel Xeon E5-2699 v4
(2.2 GHz, 2x 22core)
105,690 52,952 2,402 1,203 Red Hat 7.2
JDK 8u72
Lenovo Flex System x240 M5
2 x Intel Xeon E5-2699 v3
(2.3 GHz, 2x 18core)
80,889 43,654 2,247 1,213 Red Hat 6.5
JDK 8u60
Cisco UCS C220 M4
2 x Intel Xeon E5-2699 v4
(2.2 GHz, 2x 22core)
94,667 71,951 2,152 1,635 Red Hat 6.7
JDK 8u74
 

Note: under Performance, the max column contains SPECjbb2015-MultiJVM max-jOPS results, and the crit column contains SPECjbb2015-MultiJVM critical-jOPS results. Under Perf/Core, the max column contains SPECjbb2015-MultiJVM max-jOPS results divided by their respective core count, and the crit column contains SPECjbb2015-MultiJVM critical-jOPS results divided by their respective core count. The Environment column contains the operating system version, the JDK version, and any special configuration.

Configuration Summary

System Under Test:

SPARC T8-1 Server
2 x SPARC M8 processor (5.06 GHz)
512 GB memory (16 x 32 GB dimms)
Oracle Solaris 11.3 (11.3.23.5.0)
Java HotSpot 64-Bit Server VM, version 1.8.0_144
 

Benchmark Description

The benchmark description, as found at the SPEC website.

The SPECjbb2015 benchmark has been developed from the ground up to measure performance based on the latest Java application features. It is relevant to all audiences who are interested in Java server performance, including JVM vendors, hardware developers, Java application developers, researchers and members of the academic community.

Features include:

  • A usage model based on a world-wide supermarket company with an IT infrastructure that handles a mix of point-of-sale requests, online purchases and data-mining operations.
  • Both a pure throughput metric and a metric that measures critical throughput under service level agreements (SLAs) specifying response times ranging from 10ms to 100ms.
  • Support for multiple run configurations, enabling users to analyze and overcome bottlenecks at multiple layers of the system stack, including hardware, OS, JVM and application layers.
  • Exercising new Java 7 features and other important performance elements, including the latest data formats (XML), communication using compression, and messaging with security.
  • Support for virtualization and cloud environments.

See Also

Disclosure Statement

SPEC and the benchmark name SPECjbb are registered trademarks of Standard Performance Evaluation Corporation (SPEC). Results from http://www.spec.org as of 8/28/2017.

Fujitsu SPARC M12-2S 54,434 SPECjbb2015-MultiJVM max-jOPS, 34,771 SPECjbb2015-MultiJVM critical-jOPS; Cisco UCS C240 M5 179,534 SPECjbb2015-MultiJVM max-jOPS, 58,094 SPECjbb2015-MultiJVM critical-jOPS; Lenovo ThinkSystem SR650 177,561 SPECjbb2015-MultiJVM max-jOPS, 54,418 SPECjbb2015-MultiJVM critical-jOPS; Huawei FusionServer 2288H V5 175,588 SPECjbb2015-MultiJVM max-jOPS, 48,977 SPECjbb2015-MultiJVM critical-jOPS; SPARC T8-1 153,532 SPECjbb2015-MultiJVM max-jOPS, 89,980 SPECjbb2015-MultiJVM critical-jOPS; Huawei RH2288H V3 121,381 SPECjbb2015-MultiJVM max-jOPS, 38,595 SPECjbb2015-MultiJVM critical-jOPS; HP ProLiant DL360 Gen9 120,674 SPECjbb2015-MultiJVM max-jOPS, 29,013 SPECjbb2015-MultiJVM critical-jOPS; SPARC T7-1 120,603 SPECjbb2015-MultiJVM max-jOPS, 60,280 SPECjbb2015-MultiJVM critical-jOPS; HP ProLiant DL380 Gen9 105,690 SPECjbb2015-MultiJVM max-jOPS, 52,952 SPECjbb2015-MultiJVM critical-jOPS; Huawei RH2288H V3 98,673 SPECjbb2015-MultiJVM max-jOPS, 28,824 SPECjbb2015-MultiJVM critical-jOPS; Cisco UCS C220 M4 94,667 SPECjbb2015-MultiJVM max-jOPS, 71,951 SPECjbb2015-MultiJVM critical-jOPS; HP ProLiant DL380 Gen10 84,142 SPECjbb2015-MultiJVM max-jOPS, 25,431 SPECjbb2015-MultiJVM critical-jOPS; Lenovo ThinkSystem SR650 83,909 SPECjbb2015-MultiJVM max-jOPS, 26,145 SPECjbb2015-MultiJVM critical-jOPS; Lenovo Flex System x240 M5 80,889 SPECjbb2015-MultiJVM max-jOPS, 43,654 SPECjbb2015-MultiJVM critical-jOPS; SPARC T5-2 80,889 SPECjbb2015-MultiJVM max-jOPS, 37,422 SPECjbb2015-MultiJVM critical-jOPS; HP ProLiant DL380 Gen10 74,086 SPECjbb2015-MultiJVM max-jOPS, 62,592 SPECjbb2015-MultiJVM critical-jOPS; SPARC S7-2 65,790 SPECjbb2015-MultiJVM max-jOPS, 35,812 SPECjbb2015-MultiJVM critical-jOPS; IBM Power S812LC 44,883 SPECjbb2015-MultiJVM max-jOPS, 13,032 SPECjbb2015-MultiJVM critical-jOPS.

Brian Whitney

Principal Software Engineer

I have been doing performance work for over 30 years.  I have been involved with various benchmarking organizations, including SPEC.


Previous Post

Apache Spark SQL: SPARC T8-1 Up To 2x Advantage Under Load Compared to 2-Chip x86 E5-2630 v4

Brian Whitney | 3 min read

Next Post


Join Us at Oracle Open World 2017!

Cindy Swearingen | 1 min read
Oracle Chatbot
Disconnected