FSX and Tom's Hardware

[Guest SuperFortress]

Can anyone tell me why Tom's Hardware is constantly posting different results for their video card benchmarks in FSX. I recall their first article several months ago showed the 8800GTS as being the fastest. Then another article showed up which indicated the 280 GTX was the fastest. Today there's a brand new article which comes full circle and once again states that the 8800GTS is the winner for FSX. Very confusing.http://www.tomshardware.com/reviews/radeon...on,2007-18.html

[tfm 169]

Presumably they re-run the tests from time to time with the latest drivers.Tim

[Guest JIMJAM]

Nice pretty graphs and worthless.Read the bottom post and you get a idea of how much weight readers are given them.In Tom's defensive though, Nvidia changes drivers like diapers and the ATI 4870 X2 is not even listed. I guess it was not available to them and also neither are the latest ATI drivers.So if you are a Nvidia fan you will at least get all warm and fuzzy feeling but otherwise, IMO a waste of time.Ive owned both Nvidia and ATI and all that seems to matter is fps.Fps is like the ultimate judge of if a card is crap.Well FSX is not Unreal tournament. Stability and picture quality are just a important if not more so to me.Give it a couple weeks and all will change again.

[Guest D17S]

'pears to me this just continues to describe that anything beyond an 8800GT is not used by FS. Once an adequate conduit is in place, the CPU becomes the limit. That tight 30FPS grouping is telling us the Vcard was not limiting anything and CPU was maxed out. Sure hope it wasn't a stuttering, choppy mess. We might be grateful (at least) they didn't say it was. Vista SP1 dropped frames by 50%? Don't think so. Sounds like Tom's has system issues. In any case, we all are finally becoming aware that raw FPS numbers are meaningless for FSX evaluations. Sure wish they'd chart their Specific settings. Just "Ultra Hi" won't do it. We need it line by line. Who's a member or their forum? Give 'em a heads-up

[Mace 1,902]

I often find Tom's results to be a little strange. Not that I'm knocking Tom's--it's a good site that has its uses--it's just that I think for really meaningful FSX or FS9 testing you have got to approach it differently than they are doing.To be honest I think if you hang out in this forum or some of the tech forums on the other major FS sites, you can get better testing data than you ever could off Tom's. However, I know what's going on :) --thousands of viewers see Tom's--and never see the good testing methods some of us FS-diehards use.Consequently, the average fella buys FSX, and can't config their rig properly, partly based on incomplete testing methods as on Tom's that they saw.RhettFS box: E8500 (@ 3.80 ghz), AC Freezer 7 Pro, ASUS P5E3 Premium, BFG 8800GTX 756 (nVidia 169 WHQL), 4gb DDR3 1600 Patriot Cas7 7-7-7-20 (2T), PC Power 750, WD 150gb 10000rpm Raptor, Seagate 500gb, Silverstone TJ09 case, Vista Ultimate 64ASX Client: AMD 3700+ (@ 2.6 ghz), 7800GT

[jackcnd 0]

A FSX benchmark should publish:1) Detailed system specs2) Detailed Game config and any .cfg file changes3) A User Repeatable scenarioFSX made benchmarking much easier as you can save and share setups and with FRAPS and FS recorder it should be easy to bench a repeatable scenario?

[Mace 1,902]

Yes those are good criteria. The trick would be persuading them to do that sort of more in-depth testing. I think for that, people would have to come to a dedicated FS site like this one. Unfortunately, a site like Tom's--a benchmark like theirs--is going to get more viewing than a bench posted here.RhettFS box: E8500 (@ 3.80 ghz), AC Freezer 7 Pro, ASUS P5E3 Premium, BFG 8800GTX 756 (nVidia 169 WHQL), 4gb DDR3 1600 Patriot Cas7 7-7-7-20 (2T), PC Power 750, WD 150gb 10000rpm Raptor, Seagate 500gb, Silverstone TJ09 case, Vista Ultimate 64ASX Client: AMD 3700+ (@ 2.6 ghz), 7800GT