Benchmarking and systems performance - hosted by Elisabeth Stahl
Autores do Blog: ELISABETH STAHL 0600029JQ8 is Chief Technical Strategist and Executive IT Specialist, IBM Systems and Technology Group, and has been working in systems performance for over 25 years.
estahl 0600029JQ8 Marcações:  ultrasparc t2 sun specjappserver2004 t5240 3 Comentários 4.748 Visualizações
This morning I read that for some airlines the price of one of their shares is less than the price of a beer at the airport.
And that reminded me of some pricing claims in Sun's latest press release on their new UltraSPARC T2 servers.
Besides comparing benchmark results based on different numbers of cores and vastly different amounts of memory, comparing new Sun systems with other's old systems, and comparing actual power consumption with nameplate power consumption - big benchmark no-no's - Sun uses pricing in a footnote on a SPEC benchmark. A benchmark where pricing is not a metric. Surely meant to distract us from the fact that for these results the IBM POWER6 core offers 43% more performance than the UltraSPARC T2 core.(1)
If you're at the airport reading this blog and waiting for an MD-80, it's probably time for that beer.
(1)SPECjAppServer2004 IBM Power 570 (4 cores, 2 chips, 2 cores/chip) result of 1,197.51 JOPS@Standard vs. Sun SPARC Enterprise T5240 (16 cores, 2 chip, 8 cores/chip) result of 3,331 JOPS@Standard.
Source: www.spec.org; Results as of 4/10/08.
The postings on this site solely reflect the personal views of the author and do not necessarily represent the views, positions, strategies or opinions of IBM or IBM management.
SPEC, SPECint, SPECfp, SPECjbb, SPECweb, SPECjAppServer, SPECompM, SPECompL, SPECsfs, SPECpower, and SPECpower_ssj are trademarks of the Standard Performance Evaluation Corporation (SPEC).
technorati tags: IBM, systems, performance, Sun, benchmark,SPECjAppServer2004,T2,T5240,UltraSPARC,benchmarking