Volcane
New member
Posted by Kyle 1:04 AM (CDT)
Quack2 - The Filtering:
TechReport moves forward in showing us more optimizations in 3DMark03. And quite possibly another reason why the benchmark should not be used by any journalist or company doing hardware evaluation. I don't know whether to puke, laugh, or cry.
Typical response from Kyle, blaming 3dmark03 rather than the company that made the 'optimisations'. Nevermind that these 'optimisations' could also be being done in 'GAME' benchmarks and therefore should game benchmarks not be used for hardware evaluation also.
I actually hope the splinter cell anomaly turns out to be another of nVidias 'optimisations' then he might get his head out of nVidias ass long enough to realise that if graphics cards makers had integrity and didn't do these 'optimisations' then there wouldn't be a problem with any benchmark.
Just because they can 'optimise' doesn't mean they should, but because nVidias GFFX has a poor architecture and has trouble competing with the ATi cards, they took it on themselves to make a mockery of all benchmarking, and in doing so have probably kicked themselves in the teeth, as it is going to be hard to trust any reviews or benchmarks anymore.
The only way I can see that reviews sites are going to have to do to make fair benchmarks is probably not to use any on the rails benchmarking at all, but to actually play a game, and at certain points stop and take a fps reading, repeat at maybe 10-20 other points and then work out an average of all fps results collated and repeat for all other cards. All because of nVidia having a bad card.