I don't see enough information in the benchmark to draw any conclusions, but this quote is the key issue: "Are 32 slow threads quick enough to beat two or four fast cores?" For some applications, such as a single non-parallel operation, the answer is obviously no. Where the Coolthreads technology could be as fast or faster is for highly parallel operations. Either a highly parallel batch job (as Tanel mentioned) or an OLTP database with many transactions of short duration (think call center operators pulling customer records). I haven't seen any good benchmarks for that type of application. Keith > How does this forebode for our plans to purchase a pair of T5120s for a 10g > RAC environment? > > Thanks, > Jeff > > > On Dec 6, 2007 9:56 PM, Tanel Poder <tanel.poder.003@xxxxxxx> wrote: > >> Hi, >> >> This link below is not from an Oracle benchmark, but still a database >> one... >> unless you have many parallel threads of work running (nightly batch jobs >> for example), the CPU peformance can suck... >> >> >> http://tweakers.net/reviews/649/8/database-test-sun-ultrasparc-t1-vs-amd-opt >> eron-pagina-8.html >> >> -- >> Regards, >> Tanel Poder >> http://blog.tanelpoder.com >> >> >> > -----Original Message----- >> > From: oracle-l-bounce@xxxxxxxxxxxxx >> > [mailto:oracle-l-bounce@xxxxxxxxxxxxx] On Behalf Of Lou Avrami >> > Sent: Friday, December 07, 2007 04:48 >> > To: oracle-l@xxxxxxxxxxxxx >> > Subject: Any issues with Sun Coothreads Servers? >> > >> > Hi all, >> > >> > The organization where I am working needs to buy some Sun >> > Sparc servers for a new project. Some of the server models >> > are consideration are the T2000, the T5120 and the T5220. >> > They have requested that I ask if anyone has had any negative >> > experiences with Sun's multi-core, multithreaded CoolThreads >> > servers running Oracle. >> > >> > Any feedback, good or bad, would be appreciated. >> > >> > Thanks, >> > Lou Avrami >> > -- >> > //www.freelists.org/webpage/oracle-l >> > >> > >> >> -- >> //www.freelists.org/webpage/oracle-l >> >> >> > -- //www.freelists.org/webpage/oracle-l