I'm not sure which one is making the biggest difference, but I'd have to err on it being the CPU because it is mucho overclocked and FS-anything loves CPU grunt first.The proof will be in the pudding if someone with either the same 7600 card as yours with a C2D overclocked to around 3.5GHz or with a similar 4800+ CPU as yours and 7900 GTO/GTX runs the benchmark and posts here. Even if the former combo comes in with a lower clocked C2D, I can underclock mine to make the comparison more applicable.Gary

Just a little joke about Vista. You reported something positive about it...that can lead to World War III. I hope it doesn't.:-)

Well here's mine!!CPU: C2D E6700 @ 2.67GHz (Stock)GPU: ATI X1950 crossfires (In Non Xfire mode)Memory: DDR2 PC5300, 2GB, 667mhzOS: Windows XP Media EditionFSX Benchmark: 66 (21 & 45)

Tom,Based on the PM you sent me 10 days ago, I think you have forgotten to mention some of the tweaks you have. In particular, autogen per cell quantities and whether default.xml were used will make quite a difference to the benchmark result. Also, if you are using FRAPS to get your average FPS, make sure its FPS number is reading about the same as your FSX counter. If you run dual screen, the FRAPS counter sometimes reports double the actual FSP, so you may need to halve your results. Nonetheless, a nice result for you. I just tried making some of your tweaks and got about 20 FPS spot view and 27 FPS VC. 45 FPS for the second one of yours is very nice! Gary

>Tom,>>Based on the PM you sent me 10 days ago, I think you have>forgotten to mention some of the tweaks you have. In>particular, autogen per cell quantities and whether>default.xml were used will make quite a difference to the>benchmark result.>>GaryGood point! That's going to be an issue for everyones numbers, as I'm sure most have tweaked to one point or another.Mine were to rename the default.xml file like you state, I set max cells to 600 trees, 800 buildings, my fiber frames are set to 0.43 and my Texture_bandwidth_Mult is set to 400. Also I have deleted my BufferPool section. I also did not apply the texture resize tweaks. Incidently, I just set up the same scenario in FS9 my numbers there were 97 spot and 193 VC for a total of 290. That's with max FS9 settings including AA which as the instructions state I set my video card to let application decide. So while I'm getting decent frames in FSX, it still has a long way to go to get anywhere near that of FS9. At the levels it's at now, it's acceptable for non complex aircraft both GA and airliners. It's going though to have trouble with a PMDG or LDS level aircraft especially in the tough areas like NYC.I use the counter in ATI Tray Tools, and they do match FSX. I don't have multiple monitors, I display FSX (and FS9) using a DLP projector on an 8 foot screen!! (Actually a wall)

>Well here's mine!!>>CPU: C2D E6700 @ 2.67GHz (Stock)>GPU: ATI X1950 crossfires (In Non Xfire mode)>Memory: DDR2 PC5300, 2GB, 667mhz>OS: Windows XP Media Edition>>FSX Benchmark: 66 (21 & 45)>I have to amend my numbers due to the fact, I missed the point about unpausing the sim. My new numbers areCPU: C2D E6700 @ 2.67GHz (Stock)>GPU: ATI X1950 crossfires (In Non Xfire mode)>Memory: DDR2 PC5300, 2GB, 667mhz>OS: Windows XP Media Edition>>FSX Benchmark: 54 (18 & 36)Sorry! Still not bad for FSX!One thing I noticed. The saved config file doesn't change resolutions to 1024x768 as stated, (I changed it manually) Also a benchmark should be something that is static, with the least amount of variance. I see the config settings also included traffic. As this can vary it's probably better that no traffic is set for this test. Just my opinion!!

Something strange here.I get 12 in spot and 16 in VC. I dont think my PC likes that config. Will experiment.

A report from a (today) less than high-end system - but perhaps more representative of what the average user of FSX is using :-) CPU: Intel P4 3.0 HTGPU: ATI 9800 XT 128 MBMemory: PC-3200 (DDR400) 1,5 GBOS: Win XP FSX Benchmark: 16 (6+10)Remarks: MyTrafficX, Cloud9 XClass replacement landclass and Adam Mills' terrain textures, Autogen at:TERRAIN_MAX_AUTOGEN_TREES_PER_CELL=1500 //4500 defaultTERRAIN_MAX_AUTOGEN_BUILDINGS_PER_CELL=700 //3000 default(no other Tweaks)(This was after applying the MS upgrade to DualCore/Hyperthreading CPU's mentioned in this post:http://forums.avsim.net/dcboard.php?az=sho...id=387645&page=.. before 6+9 so this may have made a difference, but may also have been just one of those things :-hmmm FPS was measured with the Average FSP tweak of the fsx.cfg file, and there was definetly an improvement on the 2nd digets.)Disabling bloom gave a more noticable improvement:FSX Benchmark: 21 (7+14)But not a surprise given the age and "generation" of the GPU (doesn't support SM 3.0).3 years sure makes a lot of difference in specs - but still I expect to wait it out for FSX SP1, and probably also the Vista FSX SP and DX10 cards becoming "mainstream" before upgrading to a new rig.Lars

CPU: AMD Athlon64 3500+ @2.21 GHz (stock)GPU: ATI X850XT 256MB (stock)Memory: 2GB PC3200 DDR (stock)OS: Windows XPFSX Benchmark: 27 (15 + 12)Remarks: XClass replacement landclass and Adam Mills' terrain textures

TERRAIN_MAX_AUTOGEN_TREES_PER_CELL=1500 //4500 defaultTERRAIN_MAX_AUTOGEN_BUILDINGS_PER_CELL=700 //3000 default(no other Tweaks)If you haven't done so yet, rename your default.xml file in your autogen folder. It is one of the biggest bottlenecks affecting performance. It was the same in FS9. You won't see much detail difference, and it will still be much greater than a maxed out FS9, but you'll see a significant frame boost!!

>CPU: AMD Athlon64 3500+ @2.21 GHz (stock)>GPU: ATI X850XT 256MB (stock)>Memory: 2GB PC3200 DDR (stock)>OS: Windows XP>>FSX Benchmark: 27 (15 + 12)>>Remarks: XClass replacement landclass and Adam Mills' terrain>texturesAre those numbers reversed? Everyone has been reporting greater frames in VC than spot, (Something that interesting to begin with, since I generally get better numbers in spot, than VC) your's is the first to report the opposite. I'm curious, why the difference on your system??

Yes, I transposed the numbers. They should read 27 (12 + 15).

I just did one other test. Ran the benchmark, but used my current config settings, which is similar to the ones used but no light bloom or lens flare, my scenery is extremely dense Autogen dense, weather is maxed Airliner AI 100% all other off. I also handle AA & AF on my graphic and driver both at 8X. I also turned on Xfire mode! With this config, I get37 spot, 45 VC for a benchmark of 83I'm sure the difference is no light bloom and the AA and AF being handled by the video card!!It may be a good idea to have people run their current settings also, if anything, it will tell them if they can improve on what they are currently getting with their setup.

@Aerdt: Congratulations! I find it amusing and odd that this is going over as well as it has, because this exact thing was attempted in October, and it was universally poo-poo'ed. :) It was said, "it will be non-standard" and "unscientific" etc. etc.RESET_MCP (Gary) suggested it back then and so did I, but it is bon to see that it has gotten off the ground here.It will be useful to system builders especially...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

It is a great idea. Just have to work a little on a good settings template (AI Traffic add-on or not? etc.) so it is consistent.Then, after more results pile in, the data needs to go in tabular form somewhere so you can compare different rigs easily.