Benchmarks på UT2003 DEMO ikke til å stole på?

[EviLEddiE]

I denne artikkelen forklarer folkene i hardwarebladet MaximumPC grunnen til at de valgte å ikke ta med benchmarkresultatene fra UT2003-demoen. Et lite utdrag:

 

Quote:

With one of the beta version of the UT demo that we got, it would run differently depending on the platform. A P4 would produce one result while an Athlon would produce a different result. We tested with P4/Radeon 9700 Pro as well as Athlon 2800+/Radeon Pro and Athlon 1.33 T-Bird/GF3Ti500. Athlon results identical, but the scripts/runs were totally different than the P4. All P4 results were the same. I'm not really sure how web sites obtained their results when using beta UT2003 on cross-platform tests, but it's quite possible their beta versions were different than ours and perhaps the problem only cropped up on ours. We only dropped the *beta* version from our Athlon vs P4 test because it wasn't very useful.

 

Kort sagt gir benchmarken samme resultater innenfor samme plattform (AMD/Intel).

 

Men MaximumPC understreker at de ikke har opplevd slike problemer med benchmarken i det ferdige spillet.

 

Vel, jeg testet det selv. Fikk samme resultat på Flyby (149fps), men fikk faktisk bedre resultat på Botmatch (61fps i retail mot 56fps i demo).

 

Men så skal det sies at [H]ardOCP sitt testskript er mer til å stole på ved benchmarking.