T2000 Try and buy: Customers see value

Tuesday Jul 10, 2007

“With Sun’s Try and Buy program, it was easy to try a Sun Fire T2000 server—and the result was a 50 percent improvement in key customer service processing times.”
— Randy Mills, Supervisor of Information Systems, Benton Public Utility District

Sun's Try and Buy program allows customer to see the same world record performance and other advantages of the Sun Fire T2000 as Sun shows in benchmarks. Yes measured data from Sun, measured data on their code. (Sun doesn't point to peaks like other vendors).

You can read more about the Benton experience at: http://www.sun.com/customers/servers/benton.xml

At Sun we really like measured data on our systems, that is what we show in benchmarks, measured performance, measured wattages, etc. Then using this data you can justify actually trying your application/workload at your location.

Postscript for those that read the IBM blog

IBM really twists things around, when reading Official IBM benchmark blog http://www-03.ibm.com/developerworks/blogs/, IBM said:

Even the IBM p5-550Q result from 2005 surpasses the Sun T2000 in Lotus NotesBench D7 R6iNotes. Sun's T2000 result has 1.5 times worse price/performance and achieves only 5.5% more performance with double the cores and 2.6 times more memory than the IBM System x3650.
Did you see IBM 'bait and switch' (it was quick), they almost made the p550Q look 1.5x $/perf and 6% more performance.

Let's break it down. 8-core IBM p550Q is 2-chips and more expensive (when you configure the same amount of memory). Even in this benchmark IBM p550Q was 12% worse price/performance, and the Sun Fire T2000 was only 3% slower. Sun configured this with 64GB just because we could. If you really want to see pricing comparisons, take a look at IBM prices for 64GB in in IBM p550Q or IBM x3650. T2000 requires no more memory than any other processor implementation.

The Sun Fire T2000 Server consumed 2.5x less power than the IBM p550Q, and only half the space of the IBM system, adding up to 5X higher SWaP.

IBM even continues to avoid web benchmarks on POWER5 & POWER6, why????

Note: this isn't an official benchmark blog, I just enjoy getting information out anonymously.

DISCLOSURES (note that IBM typically doesn't list all of this required info, why?)

NotesBench R6iNotes Sun Fire T2000(1chip, 8cores/chip@1.4GHz UltraSPARC T1, 4threads/core, 64GB), 4 partitions, Solaris[TM] 10 U3, Lotus[R] Domino 7.0.2, 23200 users, $4.48per user, 19518 NotesMark tpm, 692ms avg rt. , IBM eServer 550Q, 8x1.5GHz POWER5, 32GB, 4 partitions, AIX 5L V5.3, Lotus Domino 7.0, 24000 users, $5.97 per user, 20108 NotesMark tpm, 932 ms avg rt., IBM X3650, 2x3.0GHz, Intel Xeon, 24GB, 3 partitions, SUSE9, Lotus[R] Domino 7.0.1, 22000 users, $3.47 per user, 18989 NotesMark tpm, 3056ms avg rt.,More info: www.notesbench.org as of 7/11/07.

The Sun Fire T2000 actually measured wattage during this benchmark run, IBM avoids measured wattage (why?), so we must use the following:
IBM p5 550Q power specifications calculated by applying 70% of the maximum Watts published in “Facts and Features Report”, 11/14/06, posted at ftp:/ftp.software.ibm.com/common/ssi/rep_sp/n/PSB01628USEN/PSB01628USEN.PDF