Why an FSX Benchmark?Because noone has ever done it before. A unified FSX benchmark will benefit all of us and will gives us an opportunity to get solid results with different hardware configurations. Too many times have we read "I get x fps" without knowing about the resolution, the amount of traffic, the quality settings, tweaks and hardware configurations. Let's make this a standard and compare.This benchmark has it all. Fixed and easy to use display settings and a fixed scenario that is challenging for both the CPU and GPU. It just takes about 5 minutes of your time.What does the benchmark do?Nothing really. The benchmark will place your Cessna in an area around San Bernardino, CA, USA at a height of 3,500 ft MSL with a standardized set of display, scenery and weather settings. All you have to do is to load the attached files for the settings and the flight. The flight is on autopilot to an altitude of 5,000 ft MSL. You don't have to do anything, except to read out the fps counter.Please, don't post screenshots.################################################################Instructions:1) Download the benchmark files from http://www.ioxo.net/fsx_benchmark_2007.zip (296kb)2) Read the documentation3) Post the results in this thread################################################################Pat

CPU: AMD X2 4800+ (stock)GPU: nVidia 7600GT, 256MB (overclocked)Memory: DDR-400, 2GB (stock)OS: Windows XPFSX Benchmark: 20Remarks:MyTrafficX, Various cfg tweaks (FIBER_FRAME:0.42, TEXTURE_BANDWIDTH_MULTI:400, BUFFERPOOLS: 5000000), default autogen (not renamed or tweaked)

I tried to do something like this when the demo first came out, but there was not much interest or feedback (still didn't stop people quoting baseless FPS figures!). Hope you have better luck.Before I give this a go, just a couple of qualifying questions:1. What screen res?2. Full screen or windowed?3. AA / AF settings (particularly for the video card, as well as in game)4. At what exact altitude do you want the average to be taken? Once stabilised or approaching 5000ft?5. Have you thought about using the freeware version of FRAPS to give say a 30 sec FPS average rather than relying on the Mk I eyeball?Gary

1. The screen resolution is set to 1024x768x32 in the settings file. That way everyone can participate.2. Full Screen3. AA/AF are also set in the attached settings file (FSX controlled AA/AF). It should be also set to "Application Controlled" in the ATi/nVidia GPU settings. Or marked in the remark section.4. My fps don't really fluctuate too much, just about 4-5 fps give and take. The average should be established within the first 90 seconds during the climb to 5,000 ft.5. I haven't thought about it, but I have thought about maybe developing something to poll the fps and make the benchmark more accurate and automated. However, I don't think that it is that challenging for the average user to eyeball an average.As for myself, it was 8fps in the Spot View and about 12fps in the VC.Thanks for pointing that out. I'll add that to the documentation.

No worries, I'll use those settings then. Well, I have just run the benchmark and get the following:CPU: C2D E6400 @ 3.5GHzGPU: 7900GTO @ 650/800MHzMemory: DDR2 PC6400, 3GB, 438MHz busOS: Windows XP ProFSX Benchmark: 37 (14 & 23)Remarks: Ultimate Traffic 2004 ported over to FSX, no other tweaks. I used FRAPS for a 30 second average FPS.Gary

>2. Full ScreenFull screen? Why?RhettAMD 3700+ (@2310 mhz), eVGA 7800GT 256 (Guru3D 93.71), ASUS A8N-E, PC Power 510 SLI, 2 GB Corsair XMS 2.5-3-3-8 (1T), WD 250 gig 7200 rpm SATA2, CoolerMaster Praetorian case

Thanks, Gary. Very interesting. Our CPUs score about the same in 3D Mark (CPU), but your GPU scored about twice as much in 3D Mark (GPU), which tells me that the GPU absolutely plays a more significant role than I expected when it comes to FSX performance.Pat

Because somehow we have to agree to an FSX standard. If you want to run windowed, you can do that, but put it into the remarks.Thanks Rhett!

I concur - consistency of results is the main thing we should be after. Interestingly, when windowed versus full screen I get 3 FPS more in spot view and 3 FPS less in VC, giving the same overall score.Another note is to make sure people turn on automixture for the benchmark, otherwise the little cessna struggles all the way to 5000ft!Gary

>Thanks, Gary. Very interesting. Our CPUs score about the same>in 3D Mark (CPU), but your GPU scored about twice as much in>3D Mark (GPU), which tells me that the GPU absolutely plays a>more significant role than I expected when it comes to FSX>performance.>>PatActually, he has way overclocked his C2D (3.5 ghz). That is probably making a big difference.

CPU: AMD X2 4800+ (stock)GPU: nVidia 7600GT, 256MB (overclocked)Memory: DDR-400, 2GB (stock)OS: Windows Vista Ultimate 64FSX Benchmark: 28 (10 + 18)Remarks:MyTrafficX, Various cfg tweaks (FIBER_FRAME:0.42, TEXTURE_BANDWIDTH_MULTI:400, BUFFERPOOLS: 5000000), default autogen (not renamed or tweaked)I am surprised! Vista scores about 8 points higher on my system. Same configuration, same everything. Interesting.Update: That was not entirely correct, because MyTraffic didn't load correctly.With MyTrafficX my score dips to 23 (8 + 15). But still 3 points more than WinXP.Pat

Surprised? Why are you surprised? :-outta

Well, pleasantly surprised, because the drivers for my nVidia 7600GT are still beta for Vista and not fully matured and optimized. I am using the Forceware v100 drivers.I am not complaining ;)

################################################CPU: AMD64 3500+ 90nm 2.21 GHzGPU: ATI 9600XT - 256Mb 527MHz/250MHz Catalyst 5.10Memory: DDR400 512x2 Mb 2.0/3/2/6 1T 200.9MhzOS: Win XP SP2FSX Benchmark: Min 5 Max 9 Ave 6.97Remarks: Default install no tweaks no addonsCPU/Mem per CPUZFrames per FRAPS99#################################################scott s..

>>Thanks, Gary. Very interesting. Our CPUs score about the>same>>in 3D Mark (CPU), but your GPU scored about twice as much in>>3D Mark (GPU), which tells me that the GPU absolutely plays>a>>more significant role than I expected when it comes to FSX>>performance.>>>>Pat>>Actually, he has way overclocked his C2D (3.5 ghz). That is>probably making a big difference.Yeah, that [email protected] will be blowing that 4800+ away........