Originally Posted by hajile http://www.anandtech.com/print/4285
btw, that benchmark was dated on the IBM side (and done on a 750 which uses a slower processor). My understanding is that the real numbers for that benchmark favor POWER7 by 25+%. In addition, POWER7 doesn't suffer from scaling issues like Xeon when you begin looking at systems with many more sockets. Also noteworthy that the POWER7 design is a much larger process and the design is a couple of years older.
That article is pretty dated in general. To quote the article: "There is no doubt here: the Westmere-EX Xeon delivers with 30% higher performance than the previous x86 quad CPU record. The 40-core, 80-thread quad Xeon server can not beat the 32-core, 128-thread IBM Power 750, the RISC champion; however, the high-end IBM servers start at $100,000, two to three times more than a comparable Xeon system."
Couple things to note:
1. The E7-4870 is a Westmere core. The P7 750 is one of the first Power 7 processors that came out. So yeah, outdated benchmark is outdated.
2. Power 7 processors have 4 threads per core. If the application is written properly, there is no way that any Xeon processor can beat it since Hyperthreading is only 2 threads per core. The fact that it even kept up somewhat should speak volumes about Intel's firm grasp as the single-thread performance king.
3. IBM platforms are expensive. The article states this, and I know this from personal experience. My old company used to spec Oracle servers for our clients. For every client, we put together an Oracle server spec on Linux, and one on Power, then ask our client to pick one themselves. They perform similarly based on our benchmarks, but the Power system is always more expensive. So expensive to the point that we only see current clients who are already on Power pick the Power platform; new clients always pick the Linux platform.