Oracle, Why Even Bother?
Yesterday, Manhattan celebrated the 200th birthday of its street grid. You know, the way you can easily get from 1st Avenue and 30th Street to Madison and 55th. And of course, Fifth Avenue is the undisputed divider of the city’s famous street grid: east of Fifth is East, and west of Fifth is West. Yet there is now an argument going on about where East really begins and where West really begins. And all I ask is Why?
Oracle just published a new TPC-H BI benchmark result yesterday that was terribly unimpressive. And it just makes you wonder, Why even bother?
Why did Oracle even bother to publish a result like their new TPC-H 3TB result(1)? Why is the scalability so uninspiring – especially if you factor in processor speed? Why isn’t Oracle highlighting that the server used for this benchmark isn’t even available until September 2011? Why is Oracle comparing this performance to an IBM result from 2009? Why weren’t they able to use more threads, only one per core here? Why is the price performance so bad, 16% GREATER cost than their previous M9000 result from October?
I won’t even ask why Oracle just pulled the plug yesterday officially on Itanium support. Because that we can surely figure out for ourselves.
(1) IBM Power 595 (32 chips, 64 cores, 128 threads) TPC-H result at 3TB (156,537 Queries per hour QphH@3000GB, $20.60/QphH, configuration available 11/24/09) vs. Oracle Sun M9000 (32 chips, 128 cores, 256 threads) TPC-H result at 3TB (198,908 Queries per hour QphH@3000GB, $16.58/QphH, configuration available 4/5/11) vs. Oracle Sun SPARC Enterprise M9000 server (64/256/256): 386,478 QphH@3000GB, $19.25/QphH@3000GB, available 9/22/11.
Source: http://www.tpc.org. Results current as of 3/23/11.
TPC-C ,TPC-H, and TPC-E are trademarks of the Transaction Performance Processing Council (TPPC).
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.