Post benchmarks for FSX here. State your system specs, clocks, voltages and any other info etc.Here is a basic list of CPUs we want to compare:Intel:i3 series stock & OCdi5 6xx stock & OCdi5 7xx Stock & OCdi7 8xx Stock & OCdi7 930 stock & OCd i7 950 stock & OCdi7 970 Stock & OCdi7 975 Extreme Stock & OCdi7 980x Stock & OCd. K model core iX CPUs stock & OCd.AMD:Phenom X2 seriesPhenom X4 seriesPhenom X6 series all including black & standard editions & OCd & stock. H55 & P55 also requested. Please use FSXmark and fraps. I believe this can be found on avsim. Please disable all addons scenery etc for benchmark. Set all display sliders at medium high in FSX settings menu. Disable nhancer profiles/graphics profiles. Disable all fsx tweaks. Make sure AA & AF are chosen in FSX graphics settings. Make sure all graphic card control panel options for FSX are set to application controlled. Use resolution of 1280x1024. Note: to admin; please pin this.

Is there a certain benchmark program to use or exactly what benchmarks within FSX are you looking for? FPS will be all over the place! As an expample there are two people with the exact same hardware and the exact same OC, but one has photo scenery, PMDG MD11, FEX, MTX, etc. while the other person has no addons. The person with no addons will obviously have a higher FPS.You may want to set some parameters, like the default Baron 58 with a default paint flying over Perth, but then you do disable FTX for those of us who have it.Anyway, I am in, just let me know if there is a certain program to use.To bad about your SSD, hopefully you will be able to get a replacement soon!

Ok. I think that covers everything. Yeah. SSD was a bummer. Dunno what happened but my comp won't get past BIOS. Tried different HD bays, data cords etc. No go. I think it's kaputz. Oh well. I'll have to save up for a new one.

Isn't the point of having a "good" computer so you can run addons with good fps/smoothness? It seems of little value to know that bone stock FSX frames are at little higher with a 980x vs i5 6xxx. I would think that the better comparison would be landing at JFK with a PMDG MD-11 with 100% traffic during a thunderstorm would be the better bench test. Becuse if you can run that scenario with decent frames/smoothness, then everything else will run good too.

Good idea, but for this to have any real meaning it has to have many controls or the results will not be dependable at all, not even on the same system same flight.As an example:Test need to be run as a canned flight that can be repeated and shared and then you need to select maybe three different locations that offer a good mix - mountain terrain so that complex Mesh triangles are rendered but then we need to also include a good city scape like Manhattan so that many 3d objects are rendered as well etc etc What aircraft, what views...panned outside, VC etc?Weather is random as loaded by FSX and can impact things a great deal, so that needs to be called out (clear/overcast/ etc) and tested to a good average.AI, what are the settings, get rid of any random loading etc....Custom LOD or no, any other "custom" cfg settings etc...Now what are the menu settings? What are the driver settings? 2xSSAA 4xMSAA? 8xAF, 16? Performance LOD (+0.6) or quality (0.0 -0.2)???? Is this a vanilla SP2 or many add-ons etc.......Recording results how? Fraps log every 0.5 second? etc There is even more if you dig into it..That is how this would need to be tested if there are to be any real comparisons...

Yeah, we need to organize this in a way that comparisions make true sense. I'm not sure I would ask people to disable any addons if they don't want to, but it's very important to know:- Use FSXMark07- Use some predefined settings profile (i'd like to use something realistic, not those global high and max that come with FSXMark)- addons used (scenery, traffic...)- resolution- GPU settings- GPU driver used- overclock- post a cpu-z screenie with your hardware...any other thing I might have forgot? any suggestions on settings? I use this:http://www.divshare....ad/12970207-3b8

Good idea, but for this to have any real meaning it has to have many controls or the results will not be dependable at all, not even on the same system same flight.As an example:Test need to be run as a canned flight that can be repeated and shared and then you need to select maybe three different locations that offer a good mix - mountain terrain so that complex Mesh triangles are rendered but then we need to also include a good city scape like Manhattan so that many 3d objects are rendered as well etc etc What aircraft, what views...panned outside, VC etc?Weather is random as loaded by FSX and can impact things a great deal, so that needs to be called out (clear/overcast/ etc) and tested to a good average.AI, what are the settings, get rid of any random loading etc....Custom LOD or no, any other "custom" cfg settings etc...Now what are the menu settings? What are the driver settings? 2xSSAA 4xMSAA? 8xAF, 16? Performance LOD (+0.6) or quality (0.0 -0.2)???? Is this a vanilla SP2 or many add-ons etc.......Recording results how? Fraps log every 0.5 second? etc There is even more if you dig into it..That is how this would need to be tested if there are to be any real comparisons...

Yeah, you beat me to it.Actually we should repeat each test like 5 times, discard the highest and lowest value, and make an average of the remaining 3, or something like that, but I'm not sure people will want to go through all the hassle

Good idea, but for this to have any real meaning it has to have many controls or the results will not be dependable at all, not even on the same system same flight.As an example:Test need to be run as a canned flight that can be repeated and shared and then you need to select maybe three different locations that offer a good mix - mountain terrain so that complex Mesh triangles are rendered but then we need to also include a good city scape like Manhattan so that many 3d objects are rendered as well etc etc What aircraft, what views...panned outside, VC etc?Weather is random as loaded by FSX and can impact things a great deal, so that needs to be called out (clear/overcast/ etc) and tested to a good average.AI, what are the settings, get rid of any random loading etc....Custom LOD or no, any other "custom" cfg settings etc...Now what are the menu settings? What are the driver settings? 2xSSAA 4xMSAA? 8xAF, 16? Performance LOD (+0.6) or quality (0.0 -0.2)???? Is this a vanilla SP2 or many add-ons etc.......Recording results how? Fraps log every 0.5 second? etc There is even more if you dig into it..That is how this would need to be tested if there are to be any real comparisons...

Here's FSXMark07, very easy to usehttp://library.avsim.net/search.php?SearchTerm=fsxmark07.zip

I don't see how flying a default CRJ700 could possibly make a realistic comparison between the listed computers. I'm personally interested in how it behaves in the premium addons (PMDG MD11, 747, FSLabs Concorde) with premium scenery (FlyTampa, FSDT....).Plus, I've noticed on my PC I get roughly the same FPS (28-30 minimum) regardless of whether I fly the concorde or MD11, and whether I'm flying in FlyTampa Hong Kong, Boston, Aerosoft Heathrow or FSDreamTeam KJFK.

I'm also more interested in those planes (got them all) but the saved flight won't work with the PMDG 744 since it has a custom AP (tried it already)Any ideas on how to set up a canned flight with those planes we can all use in our tests with fraps?Anyway, I would not like to exclude anyone here, so if anyone is interested in default planes performance, I will also run the tests with them

I don't see how flying a default CRJ700 could possibly make a realistic comparison between the listed computers. I'm personally interested in how it behaves in the premium addons (PMDG MD11, 747, FSLabs Concorde) with premium scenery (FlyTampa, FSDT....).Plus, I've noticed on my PC I get roughly the same FPS (28-30 minimum) regardless of whether I fly the concorde or MD11, and whether I'm flying in FlyTampa Hong Kong, Boston, Aerosoft Heathrow or FSDreamTeam KJFK.

I don't see how flying a default CRJ700 could possibly make a realistic comparison between the listed computers. I'm personally interested in how it behaves in the premium addons (PMDG MD11, 747, FSLabs Concorde) with premium scenery (FlyTampa, FSDT....).Plus, I've noticed on my PC I get roughly the same FPS (28-30 minimum) regardless of whether I fly the concorde or MD11, and whether I'm flying in FlyTampa Hong Kong, Boston, Aerosoft Heathrow or FSDreamTeam KJFK.

Good Point Ivan.This is just the tip of the icebergI agree that for some of us we want the test to be under the most demanding situation like PMDG 747 at Dteam Kord/JFK with ManhattanX at highres high AA/AF and LOD or Orbx product running fast and low etc - These are all good test for the extreme end and which must be pointed out - Separates the better hardware from the fair/good hardware (new build planning ahead for add-ons).That’s why things have to be opened up a bit as well as the target needs to be defined.You want to go after GPU performance then test using filtering (AA/AF) as well as bloom and shadows.You want to go after CPU and memory subsystem then you need to turn off filtering as well as vsync and any other caps.Or you can go after the overall balance of the whole with Vsync/AA/AF etc on and off - but just realize that it gets hard to pinpoint what part of the equation is performing. etc etc

Good ideas all round. However, regarding premium add-ons, I don't think the idea is to see how hardware affects these. I think we should keep it simple to see the relative effects of changes in hardware.I tried to make a start: https://spreadsheets...uthkey=CKrfrI8DUseful thread: http://www.flightsim...howtopic=140426Here's the zipfile we used: http://library.avsim...m=fsxmark07.zipLet me know if you want me to edit the spreadsheet for you guys, I'd be happy to help.

@Bangoman: Here's my fraps results with the Max settings. I couldn't do the "global" because for some reason it crashes my FSX every time before it even starts loading. Frames Time (ms) Min Max Avg17348 300000 33 79 57.827

You want to go after GPU performance then test using filtering (AA/AF) as well as bloom and shadows.You want to go after CPU and memory subsystem then you need to turn off filtering as well as vsync and any other caps.Or you can go after the overall balance of the whole with Vsync/AA/AF etc on and off - but just realize that it gets hard to pinpoint what part of the equation is performing. etc etc

To use this to test GPUs in FSX we would need settings that somehow get FSX to be GFX bottlenecked. Even if I managed to do that with my 460 someone with a 480 and an I5 750 could still be limited by his CPU.Well, that's assuming FSX is really so CPU bound, so those tests would be also interesting. I tried disabling AA & AF and a resolution 1280x720 and still couldn't see any difference when running the tests with the 460 @ stock vs overclocked@Ivan, those results at maxed settings? I was getting some 10FPS

Apologies, I mixed the two up. Those were on GLOBAL settings, the MAX settings are those that crash my sim :)By the way, I was doing the test at 1920x1080 with x8 Supersample AA and 16x AF, with x4 Transparency AA.Aaand, here are the MAX SETTINGS results as well.It comes as no surprise that they are terrible: Frames Time (ms) Min Max Avg3367 300000 ..............3.... 20...11.223

To use this to test GPUs in FSX we would need settings that somehow get FSX to be GFX bottlenecked. Even if I managed to do that with my 460 someone with a 480 and an I5 750 could still be limited by his CPU.Well, that's assuming FSX is really so CPU bound, so those tests would be also interesting. I tried disabling AA & AF and a resolution 1280x720 and still couldn't see any difference when running the tests with the 460 @ stock vs overclocked@Ivan, those results at maxed settings? I was getting some 10FPS

Why would you turn off filtering to test the GPU or ANY diff with O/C the GPU?It simple and was already pointed out to "Bottleneck the GPU" simply use all the shader effects, use Bloom, use all shadows (include terrain) us full water or at least high, use highres and high filtering.Why on earth would youexpect to see the difference in O/C the GPU while running a low res and no filtering? that would be for taking the GPU out of the mix.You have to turn off filtering in game and in driver, also disable vsync and use unlimited frames to fully test CPU and memory vs CPU and memory systems/settings.For GPU it is reverse except leave vsyc off for half of the test and then with it on watch for lowest frame...etc etc