Yes, as it happens, it's called pricewatch.com.Originally Posted by [b
As for benchmarks, just do a search. I have come to distrust them, though, as everybody manages to fudge their tests in favor of their preferred OS. I don't think much of any benchmark test that isn't performed on identical machines, varying by no more than one part (or type of part, in the case of things like arrays of RAM) or piece of software. Apple always publishes the benchmarks run with number-crunching science apps and specially tuned versions of Photoshop, because they benefit greatly from the dual-processing of which Apple is so fond, as well as the special vector processing commands in the G4 and G5 chipsets. Windows types like to look at things like games, which don't often benefit from Apple's special enhancements and - I suspect - run slower in general on Macs as a result of being ports which weren't written with the Mac OS in mind. (For example, some DirectX games utilize hardware acceleration that simply isn't available on Macs because there is no DirectX for Mac.) I think that, in general, the only valid way to compare two systems with so little in common as a Windows and Mac machine is to perform precisely identical operations, such as adding long columns of numbers. Otherwise, you can always argue that one machine underperformed due to ineffcient or otherwise differing code (or hardware, I suppose.) The interaction of the OS, the hardware, and the benchmark applications is too complex to make many objective judgements; you can get notably different benchmark results on identical systems for reasons as benign as time of day and variance in user input.
But maybe you're talking about benchmarking components, and not systems, in which case Xbench is the way to go for Mac stuff, so far as I'm told. But if I feel compelled to examine my system's performance, I like to use BLAST, GIMP or my old programming projects, because they're more fun to play with than something that just draws lots of colored squares on the screen.