Why are benchmark tests that are
conducted by the heavy hitters like Tom’s
Hardware, ARStechnica,
and others so insistent that we all can understand them and find
them useful? Who the hell has a copy of Quake III, or what-ever-in-the-hell?
How can we possibly make them understand that in order for a benchmark
to be useful, they should use the common junk that the common slob
runs. Word, Netscape, I.E. Hell… even defrag, startup, and shut
down! Forget about Winstones, whetstones, and stick-it-up-yer-ass-stones…
Just give me seconds or even milliseconds. For example:
Our test
motherboard/CPU is an ABIT KA7 with a AMD Athlon 750 processor.
It has a 10 gig hard drive and 64 megs of RAM. It has all of the
common junk that the common slob runs. Here’s how it did:
| Netscape: | 3.7 Seconds |
| I.E: | 4.1 Seconds |
| Scandisk (10 gigs): | 4 Minutes 42 seconds |
| Defrag (10 gigs): | 57 minutes 42 seconds |
| Boot up: | 1 minute 47 seconds |
I know they are controlled and scientific.
I know that there are industry standard benchmark tests that are
backed by years of research and practice. But, they are nearly unattainable
to the average person. Those graphs and stuff are really sexy, but
they don’t mean squat to me. My advice to those of you looking to
gain knowledge from benchmark tests. Look for the ones that used
Sandra 2000 or earlier.
Hardware Central uses
it and is a good place for reviews if you mean to buy stuff. Download
it and run it on your own jalopy of a computer so that you have
a realistic comparison. Don’t trust Ziff Davis stuff and their benchmarks…
they have a firm commitment to their sponsors not to piss them off.
There is no freaking way they are objective, unbiased reviews and
benchmarks.