Just got Acceleration and SP2, but before and after I installed it I ran the 5 minute dynamic FSX benchmark I created to get a measure of what SP2 has done for us in the FPS department. For those of you who have not seen my benchmark before, it is basically a scripted flight using the stock CRJ700 starting at the north of the city of Seattle, overflies the city center, overflies KSEA then heads out towards Mt Rainier. As such, it gets pretty good coverage of all scenery types in FSX, including water, cityscape, urbanscape, airport and all types of AI. I use FRAPS to record the FPS each second of flight and the max/min average. I use the default settings FSX detects for my system, which is global high across the board, plus I enable in game AA/AF and run windowed at 1680x1050.I normally use this benchmark to test hardware changes and hence post the results in the hardware forum, but this one is purely an SP1 to SP2 comparison, so here I am in the big FSX playpen!Anyway, the results of the run were:http://forums.avsim.net/user_files/179971.jpgFPS stats for the two runs were as follows:- SP1 SP2Min 12 13Max 35 37Avg 22.2 23.5This translates to a 5.8% boost in SP2 performance over SP1 which, while small, at least goes in the right direction. Also, there is obviously much more to SP2 and Acceleration that just performance.Next week I get Vista 64 installed and will do an RTM / SP1 / SP2 & DX9/DX10 comparison using this same benchmark if anyone is interested. Gary

>Max 15 37This doesn't agree with your graph. Looks like max for SP1 is closer to 34.Michael J.http://img142.imageshack.us/img142/9320/apollo17vf7.jpg

Thanks, that's an interesting benchmark. I haven't purchased Acceleration yet. And also, I'd be interested in your Vista64 tests also.Cheers,Pierre

A transcription error. Now fixed.Gary

Very interested.What is interesting is that at one stage SP1 is better than SP2 what were the conditions for this to happen.Re WOZ

The bit where SP1 is ahead is just after the dip passing Boeing field up to passing over KSEA. This is where all the KSEA airport scenery and AI starts resolving into detail.Gary

Interesting which AI because FSX SP2 isn't that compatible with the old legacy code AI (not FSX native) as was with SP1!Maybe the reason...http://i5.photobucket.com/albums/y156/awf1/sign.jpg

Is that an Excel spreadsheet?Or is there some program similar you use for it?Interesting results....but pretty much what I've been seeing...

is it possible to share with everybody this standard benchmark of yours. upload it to avsim, so we all are could be able to compare performance in FSX. lets lock the sliders somewhere to create it as standard situation.regards,tero

Yes, I have just taken the FRAPS csv output file and graphed the data with Excel.Gary

I have thought of doing that, but as clinical as the benchmark appears, it is very difficult to get standardised results from anything other than pristine installations of FSX and the two SPs. Even in the case I have reported on above, my FSX installation has been tainted by Ultimate Traffic AI aircraft and flight schedules, X graphics clouds and FS Genesis terrain mesh and landclass. I also have WideViewAspect=True, which gives me better widescreen views, but comes at an FPS penalty compare to it not being enabled. My screen resolution is also 1680x1050, which is different from what most people use I expect.As such, it is very difficult to compare different systems, unless they are set up in exactly the same way. That's not to say it can't be done, but it takes a lot of effort to ensure equivalence of all settings. Where the benchmark is most useful is with a pristine installation and for comparing changes on the same system ie. a hardware change or an SP installation as I have done above.If this is still considered useful with the above caveats, I will put together an upload package of the benchmark for the AVSIM library.Gary

I'd love to see it uploaded Gary. Not so much to compare different systems, but I'd like to see the difference on my system between RTM, SP1, and Acceleration with SP2.Doug

Alrighty then. Give me a few days, as I want to include a reasonably comprehensive set of instructions such that consistent and comparable results are achieved.Gary

hello, a comparission on Vista between RTM - SP1 - SP2 DX9 and SP2 DX10 with your benchmark really sound nice. The problem I see is that on Vista you are punished when running in windowed mode. You should try to run full screen mode. Or you can try to fix it by hiding your startbar (autohide).Jan

Gary, thanks. As someone has already noted it will be very useful in each individual case on every PC/SetUp, since, for example, there are many add-on installations to benchmark on the same system. Besides your graphs and numbers give a good general idea of what to expect. It will be very useful and pls describe as many detials/settings as you reasonably can in the readme file. :)Thanks.