I will have a fairly comprehensive comparison using FS2k2 of my GF3TI200 64mb card that was O/C to 230/500 core/mem vs. an MSI/PNY TI4400 128mb card O/C to 300/630 ($229@BB) in a few days.Yeah, the 3dmark2001se score shot up to 12060 (link below to compare)and that was at just 1.8+Ghz! But what will she do in FS2k2? And what will an upgrade like this do for you? Should you wait for the next gen cards? Would you be satisfied with the results? Probably depends on your expectations, but I think the benchmarks will speak for themselves. :-) :-( :-) :-( :)The benchmarks are complete as far as the GF3 goes and the GF4 will soon follow then I will try to put it all together so folks can make head or tail of it. Testing was done using a canned 3 minute flight around Las Vegas with ALL sliders to max including Mesh at 100% but no ATC for obvious reasons and the latest build of Fraps was used to record FPS performance and will include testing of the following settings: @1280x960x32no AA no AF (aniso filtering)AA x2 no AFNo AA 32tap AF2x AA 32tap AF4x AA 32tap AF@1600x1200x32no AA no AF4x AA 32tap AF

This is great information. I can't wait for the final review!Robbie

Paul,Great job as this is precisely what I have been after ever since the Geforce 4 came out. Two things though. It would be interesting (perhaps important is a better word) to update the comparison when the det 5's come out, given the supposed improvement in Geforce 4 aniso level 4 and 8 performance. One would hope the Geforce 4 would gain more from this newer driver than a Geforce 3, given the theoretically close performance between the two at present with level 4 and 8 aniso on.Secondly, why not just test them at standard speed? Am I the only one around here who runs everything at stock speed? :-)

>Two things though. It would be >interesting (perhaps important is a better word) to update >the comparison when the det 5's come out, given the supposed >improvement in Geforce 4 aniso level 4 and 8 performance. >One would hope the Geforce 4 would gain more from this newer >driver than a Geforce 3, given the theoretically close >performance between the two at present with level 4 and 8 >aniso on.>Secondly, why not just test them at standard speed? Am I the >only one around here who runs everything at stock speed? :-) Jon,Those are all great suggestions, As I get better at displaying the results (there are a thousand ways) I might slap in the GF3-200 and re-run just for that purpose, I was just eager to put up some numbers and the GF3TI was running so well at those speeds that I felt "That

Great Paul - I'm really looking forward to your findings. I guess your findings will be quite instrumental in my decision to get a 4400 for $500 Australian dollars - or wait for the NV30 which no doubt will be the best part of $1000.

>Great Paul - I'm really looking forward to your findings. I >guess your findings will be quite instrumental in my >decision to get a 4400 for $500 Australian dollars - or wait >for the NV30 which no doubt will be the best part of $1000. Sorry, I'm a bit behind but its coming though. :)You really might want to consider the 4200 as you can run them at or above 4400 stock speeds, here in the states you can get the Visiontek GF4TI4200-128mb card for $179-US at BestBuy. I picked up the Gameward 4400 because it was offered to me at a special price from a freinds store as it is the older XP700 model, the new one is the XP750 and comes with faster memory, Gameward's use quality parts that will contribute to better image quality it includes the dual-head display bla bla bla But right now the XP700 I have is running at 305(core) and 680(mem) and thats not too shabby! :)Anyhow Yeah, the NV30 is not going to be cheap at least not at first. :)Later Jon,