Check the results here:http://www6.tomshardware.com/graphic/20031...38-nv36-40.htmlRegardsJaime CuellarMonterrey, MexicoMy Simulator system:PC # 1P4 2.8 GHZ 800 FSB HTMotherboard MSI NEO 465PE1 GB PC3500 DDR Corsair XMS RAMHard Drive 80 MB 7200 RPM IDEATI RADEON 9800 PRO 128Viedeo Projector Proxima LX1 100" ScreenFlight Simualaror 2004 "Century of Flight"Windows XP Pro SP1.PC # 2P4 1.7 GHZ 768 MB RAMBUS PC-800GeForce 2 GTS Pro 32 MBGeForce 2 MMX 64 MBMonitor # 1 Sony 17" FlatMonitor # 2 Acerview 15"Project Magenta (PFD & EICAS)Windows XP Pro SP1.PC # 3AMD Athlon 1300 XP 512 MB SDRAM PC133GeForce 2 MMX 64 MBMonitor HP 15"Project Magenta (CDU & MCP)Windows XP Pro SP1PC # 4Toshiba Satellite P3 Celeron 1.1 GHZ 256 RAMFsMeteo 6.01, Active Sky 1.9, PM Sounds, FliteStar 8.1 & Jeppview(Jeppesen), SquawkBox 2.3 and Roger Wilco.All PC's are networked with a registered FSIUPC 3.11 & Wide FS.Additional hardware:PFC Jetliner Yoke & PedalsSafeline Jet throttle consoleSaitek Cyborge force feedback JoystickAerosoft MCP

Go the Radeons! :-hah

John, is that for real? That is a monstrous improvement to go 9800 XT from 9700 Pro. Do you have personal experience to comfirm that? In many other benchmarks you are lucky to see 25% improvement if I remember correctly. At 4x/8x 1600x it appears to be a 70% improvement in frames. Noel

Noel,Tends to follow what I have seen when benchmarking my 9700Pro vs my 9800Pro following the same benchmark settings. What I still find very impressive is the showing that the 9700 still puts in. Lends credence to the statement that it was and probably will be one of the best hardware releases of all time.It's also nice to see the Radeons standing as king of the hill but we all know what that means. Nothing. Still nice to see though. :)Bobby

Those are some pretty amazing numbers. Despite what we've heard about FS being CPU dependent, it's obvious that the right card can yield some major improvements in FS2004.I wish they would run the same cards and benchmarks on a middle-of-the-road system, so that we can get a better sense of where the bottlenecks might occur. I have a P4 2.4 (with a main board and memory that is over 2 years old), and I'd suspect that I might not see a big jump if I swapped out my Radeon 9500 Pro for something like the 9800 XT, because the new card would be limited by my CPU. On their testbed (a P4 3.2), there is probably enough CPU headroom to let the video card really stretch its legs.

Thanks for the link.

You'll see some increase going from a 9500 Pro to a 9800XT, but nowhere near as big as what that Tom's article says. It's bogus in its comparisons between the 9700 Pro and resolution changes. I can tell you from experience that there's very little difference in framerate going from 1024x768, 1280x1024 and 1600x1200 resolutions with a 9700 Pro. Very little. Maybe 2 frames per second with each resolution increase, max; running 4xAA & 16xAF. Also, there's only about a 10 to 15 percent increase in performance between the 9800 Pro cards from the 9700 Pro cards, if that; with the latest drivers. In other words, what's being reported is total Bull. Makes me question the rest of the results too. I know that if you look straight up into the sky or straight down onto an open field from an outside view, the 9700 Pro will get you about 110 fps while the 9800 Pro will get you around 150 fps and that's running 4xAA & 16xAF. Problem is, that's not a true test of what each card will do overall in normal situations. The 9800XT gives a small boost in framerates over the 9800 Pro. Not even enough to justify the extra expense unless you have to have the latest and greatest. My suggestion, stay with the 9500 Pro if it's doing the job for you. If you do upgrade, upgrade the processor too. It'd probably be best to wait awhile. Next spring the R420 will be available and I'm sure we'll see some significant improvements in processors by then too.If what was being reported by Tom's was anywhere near accurate, I'd be running at a lower resolution or running out and buying a 9800 Pro or XT card tommorrow. I'm running at 1600x1200x32 res and getting extremely good framerates in all situations with all my graphics sliders maxed out. Virtually no difference in frames when I lower the resolution and I've tried em' all. Course, I'm running with the autogen fixes and using Chris' clouds (which are better than the originals IMHO) and have no frame problems even with heavy storms and maxed out clouds (density on high; visibility at 90mi and coverage and 3dClouds maxed out).

For some reason, Tom's Hardware has moved away from including the non-pro versions of cards in the video card comparisons. For people who can sacrifice a certain degree of performance to save a substantial amount of money (usually...in most of the classes of cards), the non-pro (or non-ultra) versions are the way to go. When I was in the market for a card a couple of months ago, I had to research a few sites to find that the 9800 (that I wound up buying) was a great bargain for a card that sat above the 9700 pro (and all 5600's and the 5900NP) and below the 9800 pro. Things have changed since then (Duh!), but my point is that Tom should include all of the options in his card comparisons. Just my opinion....JD

Plenty of people have issue with toms hardware. I usually try to go back and forth between toms and hardocp.com to get a rounded review of something.For you folks that claim to run your msfs04 with maxed out settings, I'm wondering if you really are maxing out all the settings or just some? I know its easy to say "I maxed them all out". ALSO, what do you have your control panel settings at as far as vsync, AA/AF, texture preference, mip map detail. I get very good performance on my p4 3.0C, 9800 pro 128, 1024 ddr400 ram, system but I don't get it when I MAX all the settings out. I try to have it set up so that at any time I can keep my FPS at the minimum in the 20's in heavy areas.Clouds sometimes just kill it, even with Chris's clouds.

Bigshot and all: My question has to do specifically at 1600X1200. I don't seem to have any anti-aliasing capabilities with my 9500 Pro. I have seen mention of this in other threads. I want to run at 16X12, and I like a little AA, at least I liked it when I was at the lower resolutions. I also get some pronounced slowing when on final to my home airport, KSEA, (down to about 14 fps from 22). AMD Barton 2800 with three 512 MB 2700 DDR on an AV8X. Keith

>Bigshot and all:>> My question has to do specifically at 1600X1200. I don't>seem to have any anti-aliasing capabilities with my 9500 Pro. Hi, Keith.I've got a 9500 Pro as well, and I can get anti-aliasing at 1600x1200. However, I have found that I need to reboot if I make a change at that res (particularly if my desktop is set to 1600x1200), or it won't take effect. Have you tried making the change in the ATI control panel (not in the sim itself), and then doing a reboot to make sure it sticks?I also fly out of KSEA, and it takes a bit of compromise on the sliders to keep the fps up, particularly if you are flying into 34L or 34R (with boeing field and downtown seattle in the background). I'm pretty conservative about my settings, so I manage to keep it at my locked 20fps in most situations, including Seattle--unless there is lots of clouds, in which case I will go to 1024x768. What kind of settings are you using? Just for reference, I use:4xAA and 4xAF1280x960Terrain Mesh: 90 (FSGenesis 38m)Texture size: HighTerrain Detail: Land ONlyWater effects: HighDawn/dusk blending: EnabledExtended Terrain Textures: EnabledScenery Complexity: very denseautogen: DenseSun Flare: EnabledLense Flare: disabledVC Gauge Quality: HighReflections: DisabledAircraft Cast Shadows: EnabledLanding Lights: EnabledSight Distance: 60Cloud Draw: 403d cloud percentage: 100Cloud Cover: HighRender to texture: EnabledTransform and Lighting: EnabledMip-mapping: 4Lights: 6Global max texture size: MassiveAI percentage: 75% (project AI)

What is not clear is are the FPS measured in average or is that what they saw it peak at, but it averaged lower?I have the 9500 pro and I enjoy this card tremendously, and I have my scenery set extremely high, but I average 20fps. I have it locked at 22. While I can maintain 22 in rural areas I loose about 2 frames when in the city or at busy airports like Seatac. I imagine I could get better frames too if I only used the reduced panel (W). Anyway, one thing is shown here, that Nvidia is not the Shizzen Bizzen Hizzle Dizzle anymore.....

Im picked the Radeon 9800 pro 128 card for my new graphic card last week before I even seen this thread.

Jimmy: Thanks for all of the information. I run at 1600X1200, (as I said, I can't get any antialiasing), and I use about 70% 3D cloud percentage. I'm not willing to go below 16X12, because I fly some panels with small detail. I will look at some of your settings, and try them out. I have decided to stick with my 9500, because with the exception of Tom's Hardware, all the benchmarks seem to indicate only a 10-15% difference between the 9500 and 9700, and Flight Simulator has always been known as more CPU than GPU. I'm going to try to make the 9500 work. Keith

Well, until now, I had pretty much settled on a geforce FS5950 Ultra 256MB card when I build a new pc in the next couple months. Now, I'm not so sure if I'll go FX5950 or Radeon 9800XT......