Check out this comparison of IBM Power systems to systems from Sun, Intel and HP.
According to IBM, more than 2,100 users of competitive hardware have eliminated server farms by migrating to Power Systems over the past three years. The company adds:
"IBM is the only major vendor to gain revenue share in the UNIX segment for the past five years (+11.2 points) while both Sun (-1.9 points) and HP (-5.7 points) lost share, according to IDC.
"IBM has helped nearly 1,200 customers migrate from competitive Sun, HP and other UNIX platforms to IBM's AIX or Linux on Power."
It also refutes claims from other vendors, in particular Oracle's announcement that Oracle and Sun SPARC Solaris achieved World Record TPC-C Performance beating IBM's best results on DB2 with Power 595 Server.
From IBM:
"This is not a straightforward 'apples to apples' comparison: The IBM system is a single server while the Sun system is a configuration with a 12-node cluster. The Sun cluster had a total of 384 processor cores and 3,072 threads compared to 64 cores and 128 threads in the IBM system. With 512GB of memory per node the Sun cluster had a total of 6TB of memory, compared to 4TB in the IBM Power 595 system. While the Sun tpmC per core is 20,097, the IBM tpmC per core is 95,080. The IBM result had 4.7 times higher performance. The Sun system was not available until December 2009. The IBM system had been available since December 2008.
"In late August and early September (2009), Oracle ran an advertisement in The Wall Street Journal and The Economist making unsubstantiated superior performance claims about an Oracle/Sun configuration relative to an official TPC-C result from IBM. Oracle has been fined for making claims they could not prove.
"The Transaction Processing Council (TPC) -- a non-profit corporation founded to define transaction processing and database benchmarks and to disseminate objective, verifiable TPC performance data to the industry -- has determined that Oracle's advertisement violated TPC policies by not adhering to basic standards of fidelity, candor and due diligence. Oracle, a TPC member, was fined for the violation, required to remove the advertisement from websites and report back to the TPC on their steps for corrective action and future compliance."
Read the TPC press release and download TPC's letter to Oracle.
Also, check out this YouTube video, "Sun, HP and x86 users are moving to Power." It shows how customers can use the IBM migration factory to move to Power systems. The video features a company that migrated more than 20 Sun servers to two Power 570 boxes, saving more than $500,000 annually in software costs. I imagine their electric bill dropped a bit as well.
Relatedly, Oracle and SAP customers can use the IBM insight tool to help determine how they can consolidate their current server farms to IBM Power servers.
All the information linked in this post, incidentally, was published prior to the POWER7 announcement. Of course the new systems make IBM's consolidation message even more compelling. For example, I want to highlight the rPerf numbers comparing POWER6 and POWER7 systems.
Compare a 12-core 5 GHz POWER6 (rPerf of 111.30) server to a 12-core 3.5 GHz POWER7 (rPerf of 140.75) server. Or compare the respective high-end boxes, a 32-core 4.2 GHz POWER6 570 (rPerf of 193.25) vs. a 48-core 3.5 GHz POWER7 770 (rPerf of 493.37). The numbers for the new systems are significantly better across the board.
According to IBM, more than 2,100 users of competitive hardware have eliminated server farms by migrating to Power Systems over the past three years. The company adds:
"IBM is the only major vendor to gain revenue share in the UNIX segment for the past five years (+11.2 points) while both Sun (-1.9 points) and HP (-5.7 points) lost share, according to IDC.
"IBM has helped nearly 1,200 customers migrate from competitive Sun, HP and other UNIX platforms to IBM's AIX or Linux on Power."
It also refutes claims from other vendors, in particular Oracle's announcement that Oracle and Sun SPARC Solaris achieved World Record TPC-C Performance beating IBM's best results on DB2 with Power 595 Server.
From IBM:
"This is not a straightforward 'apples to apples' comparison: The IBM system is a single server while the Sun system is a configuration with a 12-node cluster. The Sun cluster had a total of 384 processor cores and 3,072 threads compared to 64 cores and 128 threads in the IBM system. With 512GB of memory per node the Sun cluster had a total of 6TB of memory, compared to 4TB in the IBM Power 595 system. While the Sun tpmC per core is 20,097, the IBM tpmC per core is 95,080. The IBM result had 4.7 times higher performance. The Sun system was not available until December 2009. The IBM system had been available since December 2008.
"In late August and early September (2009), Oracle ran an advertisement in The Wall Street Journal and The Economist making unsubstantiated superior performance claims about an Oracle/Sun configuration relative to an official TPC-C result from IBM. Oracle has been fined for making claims they could not prove.
"The Transaction Processing Council (TPC) -- a non-profit corporation founded to define transaction processing and database benchmarks and to disseminate objective, verifiable TPC performance data to the industry -- has determined that Oracle's advertisement violated TPC policies by not adhering to basic standards of fidelity, candor and due diligence. Oracle, a TPC member, was fined for the violation, required to remove the advertisement from websites and report back to the TPC on their steps for corrective action and future compliance."
Read the TPC press release and download TPC's letter to Oracle.
Also, check out this YouTube video, "Sun, HP and x86 users are moving to Power." It shows how customers can use the IBM migration factory to move to Power systems. The video features a company that migrated more than 20 Sun servers to two Power 570 boxes, saving more than $500,000 annually in software costs. I imagine their electric bill dropped a bit as well.
Relatedly, Oracle and SAP customers can use the IBM insight tool to help determine how they can consolidate their current server farms to IBM Power servers.
All the information linked in this post, incidentally, was published prior to the POWER7 announcement. Of course the new systems make IBM's consolidation message even more compelling. For example, I want to highlight the rPerf numbers comparing POWER6 and POWER7 systems.
Compare a 12-core 5 GHz POWER6 (rPerf of 111.30) server to a 12-core 3.5 GHz POWER7 (rPerf of 140.75) server. Or compare the respective high-end boxes, a 32-core 4.2 GHz POWER6 570 (rPerf of 193.25) vs. a 48-core 3.5 GHz POWER7 770 (rPerf of 493.37). The numbers for the new systems are significantly better across the board.
Take the time to see why a customer might want to move up to Power Systems.
Note: Don't forget to check out the replay from the most recent AIX Virtual User Group meeting. The Feb. 18-19 presentation is a two-parter covering POWER7.
So, what's the deal when the source of the comparison is one of the compared players?
IBM facts will always show Power Systems are second to none.
Any independent comparison out there?
Regards
Posted by: ArlequÃn | February 24, 2010 at 07:47 AM
Amusing bit of benchmarketing there.
It was actually thanks to IBM that we had to go through a rather large and lengthy comparison of commodity x86/Linux -v- POWER6/AIX. IBM sales demonstrated that the xSeries boxes can achieve the same rPerf rating as the POWER6 systems at a fraction of the cost when sizing out a new project for us. On top of this, add on the lower IBM PVU licensing for x86 systems.....
Needless to say it raised a lot of eyebrows and made things difficult for us AIX/PowerVM fans.
I'd be curious to see exact configurations on their benchmark claims.
Posted by: Jeff | February 24, 2010 at 04:18 PM