Tuesday Apr 23, 2013

More SPARC T5 Performance Results

Performance results for the new SPARC T5 systems keep coming in...

Last week, SPEC published the most recent result for the SPECjbb013-MultiJVM benchmark. This benchmark "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" according to SPEC.

All of the published results are at: http://www.spec.org/jbb2013/results/jbb2013multijvm.html.

For the first table below, I selected all of the max-JOPS results greater than 50,000 JOPS using the most recent Java version, for the SPARC T5-2 and for competing systems. From the SPECjbb2013 data, I derived two new values, max-JOPS/chip and max-JOPS/core. The latter value compensates for the different quantity of cores used in one of the tests. Finally, the "Advantage of T5" column shows the portion by which the T5-2 cores perform better than the other systems' cores. For example, on this benchmark a 32-core T5-2 computer demonstrated 15% better per-core performance than an HP DL560p with the same number of cores.

As you can see, a SPARC T5 core is faster than an Intel Xeon core, compared against competing systems with 32 or more cores.

Model CPU Chips Cores OS max-JOPS Date Published max-JOPS per chip max-JOPS per core Advantage of T5
SPARC T5-2 SPARC T5 2 32 Solaris 11.1 75658 April 2013 37829 2364
HP ProLiant DL560p Gen8 Intel E5-4650 4 32 Windows Server 2008 R2 Enterprise 67850 April 2013 16963 2120 12%
HP ProLiant DL560p Gen8 Intel E5-4650 4 32 RHEL 6.3 66007 April 2013 16502 2063 15%
HP ProLiant DL980 G7 Intel E7-4870 8 80 RHEL 6.3 106141 April 2013 13268 1327 78%

The SPECjbb2013 benchmark also includes a performance measure called "critical-JOPS." This measurement represents the ability of a system to achieve high levels of throughput while still maintaining a short response time. The performance advantage of the T5 cores is even more pronounced.

Model CPU Chips Cores OS critical- JOPS Date Published critical- JOPS per chip critical- JOPS per core Advantage of T5
SPARC T5-2 SPARC T5 2 32 Solaris 11.1 23334 April 2013 11667 729
HP ProLiant DL560p Gen8 Intel E5-4650 4 32 Windows Server 2008 R2 Enterprise 16199 April 2013 4050 506 44%
HP ProLiant DL560p Gen8 Intel E5-4650 4 32 RHEL 6.3 18049 April 2013 4512 564 29%
HP ProLiant DL980 G7 Intel E7-4870 8 80 RHEL 6.3 23268 April 2013 2909 291 151%

As always, care should be taken in choosing a benchmark that is similar to the workload that you will run on a computer. For example, if you plan to implement a database server, using the SPECint benchmark will not help you, because that benchmark merely measures the performance of the CPU cores and speed and size of memory caches (and perhaps the memory system). It does not measure performance of network or disk I/O, and both of those are important factors in database performance - especially storage I/O.

According to the SPECjbb2013 design document, this benchmark "exercises the CPU, memory and network I/O, but not disk I/O." Because of this, it can be used as a simple method to estimate relative Java processing performance. From the data shown in the tables above, it is clear that the newest SPARC cores deliver Java performance that is competitive with the most recent Intel Xeon CPU cores.

Edit [2013.04.23]: Jim Laurent uses the same benchmark results in a quick look at the smooth scalability of Solaris 11, compared to RHEL6.

For more information on recent SPARC T5 world records, see https://blogs.oracle.com/BestPerf/.

SPEC and the benchmark name SPECjbb are registered trademarks of Standard Performance Evaluation Corporation (SPEC). Results as of 4/21/2013, see http://www.spec.org for more information.

Wednesday Apr 17, 2013

IDC Reviews New SPARC T5 and M5 Servers

On April 5, IDC published an article that provides their view of the recent announcement of Oracle T5 and M5 servers.

IDC's conclusion: "Oracle has invested deeply in improving`the performance of the T-series processors it developed following its acquisition of Sun Microsystems in 2010. It has pushed its engineering efforts to release new SPARC processor technology — providing a much more competitive general-purpose server platform. This will provide an immediate improvement for its large installed base, even as it lends momentum to a new round of competition in the Unix server marketplace."

IDC also noted the "dramatic performance gains for SPARC, with 16-core microprocessor technology based on three years of IP (intellectual property) development at Oracle, following Oracle's acquisition of Sun Microsystems Inc. in January, 2010."

The new T5 servers use SPARC T5 processor chips that offer more than double the performance of SPARC T4 chips, which were released just over a year ago. And the T4 chips, in turn, were a significant departure from all previous SPARC CMT CPUs, in that the T4 chips offered excellent performance for single-threaded workloads.

The new M5 servers use up to 32 SPARC M5 processors, each using the same "S3" SPARC cores as the T5 chips.

Wednesday Mar 27, 2013

New SPARC Servers Outrun the Competition - by Leaps and Bounds!

In case you missed yesterday's launch of Oracle's new SPARC server line, based on the SPARC T5 and M5 processors, here is a brief summary - and the obligatory links to details...

The new SPARC T5 chip uses the "S3" core which has been in the SPARC T4 generation for over a year. That core offers, among other things, 8 hardware threads, two simultaneous integer pipelines, and some other compute units as well (FP, crypto, etc.). The S3 core also includes the instructions necessary to work with the SPARC hypervisor that implements SPARC virtual machines (Oracle VM Server for SPARC, previously called Logical Domains or LDoms) including Live Migration of VMs.

However, four significant improvements have been made in the new systems:

  1. 16 cores on each T5 chip, instead of T4's 8 cores per chip, was made possible because of a die shrink (to 28 nm).
  2. An increase in clock rate to 3.6 GHz yields an immediate 20% improvement in processing over T4 systems.
  3. Increased chip scalability allows up to 8 CPU chips per system, doubling the maximum number of cores in the mid-range systems.
  4. In addition to the mid-range servers, now the high-end M5-32 also supports OVM Server for SPARC (LDoms), while maintaining the ability to use hard partitions (Dynamic Domains) in that system. (The T5-based servers (PDF) also have LDoms, just like the T4-based systems.)

The result of those is the "world's fastest microprocessor." Between the four T5 (mid-range) systems and the M5-32, this new generation of systems has already achieved 17 performance world records.

Some of the simpler comparisons that were made yesterday include (see the press release for substantiation):

  1. An Oracle T5-8 (8 CPU sockets) running Solaris has a higher SAP performanace rating than an IBM Power 780 (8 sockets) running AIX.
  2. A single, 2-socket T5-2 has three times the performance, at 13% of the cost, of two Power 770's - on a JD Edwards performance test.
  3. Two T5-2 servers have almost double the Siebel performance of two Power 750 servers - at one-fourth the price.
  4. One 8-processor T5-8 outperforms an 8-processor Power 780 - at one-seventh the cost - on the common SPECint_rate 2006 benchmark.

The new high-end SPARC system - the M5-32 - sports 192 cores (1,536 hardware threads) of compute power. It can also be packed with 32 TB (yes, terabytes!) of RAM. Put your largest DB entirely in RAM, and see how fast it goes!

Oracle has refreshed its entire SPARC server line all at once, greatly improving performance - not only compared to the previous SPARC generation, but also compared to the current generation of servers from other manufacturers.


Jeff Victor writes this blog to help you understand Oracle's Solaris and virtualization technologies.

The views expressed on this blog are my own and do not necessarily reflect the views of Oracle.


« August 2016