Dumbest question on earth

[alien]

Ok... as the title says... this is probably the dumbest question on earth... but what do benchmarks really tell us? I know that on xbench they state that it "is useful for comparing the relative speeds of two different Macintoshes", but what do the numbers actually mean? I know that "real world" performance means more than just benchmarking numbers, but I'd like to understand what the actual meaning of these numbers means.

The MacWorld article on the new PowerMacs states that the new machines are not any fast than the previous models even though the previous speedmark scores differed. The Barefeats iMac vs. dual-core PowerMac gives examples how the 2.1 ghz iMac actually beats a dual-core 2 ghz Powermac in certain tests. I just wanted to know how benchmarks actually help the consumer other than telling them that the old model had a number of "1000" and the new model has a number of "1050".

I know it's dumb, but I just would like to hear everyone's thoughts on this. Thanks.

 

[jsw]

Benchmarks tend to confuse people more than they help. It's like using the 0-60 times as your exclusive decision making tool for buying a car.

The better the number, the better the computer does on that very specific [typically non-real-world] task. Unless the numbers are substantially better (percentage-wise), you'll see little if any difference.

Also, since most users tend to use their computers for many different tasks, most of which are not CPU-intensive, benchmarks are very misleading, in the same sense that those 0-60 times don't make much of a difference when driving from Boston to LA.

Benchmarks are fun to look at but - for most people - pretty useless except when they show significant differences on tasks you'll use.

 

[Flying Llama]

Actually, I don't think this is the dumbest question on Earth. Actually, I think this is a very good question... oh yeah!

llama