Nick,My system specs are in my sig and there has been noticeable improvement with the affinity at 85. I am very pleased!

With HT on there will be a limit placed on the clock ability. In the case where users are not clocking i7 to 4-4.4Ghz that should not present a problem assuming those who may be mildly clocking have a heatsink that is able to deal with the change however those who are clocking high will need to evaluate their clocks/stability and temps to test this correctly. Some may find the loss in clock not worth the change, .. some may see improvement worth the loss and some may not, some may see no changeWhat I do not want to do is post something that only effects a certain percentage of users and have people read that and assume they should do it tooStats are needed to make that call and how it should be presented

I don't know if this is working for other corei7 users with HT enabled, but changing the mask to 85 as stated above has made a noticeable difference for me as well in FSX. I thought this was pretty strange so I disabled HT and removed the affinity mask settings in the config file and I still don't see the 5-7 FPS increase that I do with HT enabled and the affinity at 85.

It is likely that some of the operating system actions are using the hyperthreads which your 85 mask leaves free.Ideally I think having hyperthreading off and an affinity mask of 14 should be best -- real core 0 being reserved for the operating system and other ancillary programs you might be using, leaving 3 real cores for FSX. With HT enabled the equivalent mask value would be 84 to leave a real core free for Windows.If you omit the affinity mask, FSX (since SP1 I think) may be attempting to make use of all cores (or at least constantly checking on them), which is certainly a mistake with hyperthreading. I think this may be why 85 with HT gives better results than 15 or nothing. I suspect 84 (with HT) or 14 (without HT), respectively, would be better still, but I see no difference on my setup (Core i7 975 at 4.50GHz). I have HT disabled in the BIOS.RegardsPete

OK, now we are using our noodle.. Glad to see someone is applying proper logic to this instead of trial/error guess workThe result of using this technique may have value for a number of users but not all. The outcome is problematical to the system in use which incorporates everything in that system from hardware/drivers/software/OS running and possibly the scenery of choice along with influences of FSX settings and/or other edits in the config. I have 4 i7 systems here of various type proc and none of them respond any differently to a value of 85/HT ON over 15/HT OFF or the entry removed from the config completely. As a matter of fact I start noticing stutters with any affinity mask setting which restricts available physical cores which make sense since I usually run a 3/4 100% autogen slider and those cores are needed for that autogen level.I do not do silly things like run edited scenery radius values, run 4096 cloud textures, or especially edit in those autogen limiting lines in my config as they throw the priority system off and change thread use to my cores. Threads to cores other than the primary are all about terrain and autogen.I also do not use Vista or W7 for FSX. I use XP x64 and do not suffer fromi7 stuttersBlack screen of death with payware aircraftstrange menu crashesstrange colored tiles appearing on the ground or in waterCTD'snone of the common issues posted in forums all the time and I suspect many of them are related to Vista/driversAlright, so what you are dealing with and looking for is the common denominator in systems out there which with the value of 85 are in fact respecting the 4 physical cores without the thread collisions on the primary core , which is what you have managed to do here, tune that out and smooth the the screen result. What I would like to know is if the systems showing a difference are all Vista/W7 or are there any XP/x64 systems (CONFIRMED) displaying a positive change (CLUE?)HT has nothing to do with it, nor will HT in any way shape or form increase FSX performance as the application has no means of addressing virtual core threads.If I see enough people confirming a consistent result I will add this to the tuning thread under a i7 entry for JOBSCHEDULER however I can not personally confirm this because I never had any problems with smooth flight and i7 on any system I have built and set up.

Nick, Mitch,et.al.So I did set up HT on my i7940 (4.0ghz), set jobscheduler at 85, and found some added smoothness. This was confirmed over several brief flights that included one location in the US with GEX, UTX, and Caranado aircraft; and then one in the FTX modified Tamworth in the A2A J-3. The FTX modified Tamworth packs a lot of scenery textures in there (the download of this scenery, added on top of the FTX base SP3, totals 1.3 gb for the airport alone.). Without HT and 85, I was getting some black tiles for a second or two during turns. With HT and 85, those black tiles have gone away. I have some more testing to do, but I think that the improvement is greater than "placebo" level.I agree wholeheartedly with the comment that each PC is its own, and that performance will vary from unit to unit. I run Vista 64 bit on this computer that is dedicated to flight simming. I have followed Nick's guide; but have also followed his suggestions regarding the Vista "services". So I do think that allowing the hyperthreads to perform these "background" services and not steal any cycles from core 0 running FSX and core 2, 4, and 6 running the texture loads may help. I did monitor the temps during the flights (hi of 61), but have not as yet done an OCCT stress test. (will do that in the next day or two.)I enjoy tracking down some of these tips. Some have been blind alleys. Some have been cul-de-sacs. And some have led to the promised land. In any case, I feel that I am now testing flight sim/equipment in trying to move my "flight sim happiness quotient" from 9.6, on a scale of 1 to 10, to 9.9. I won't get to 10 until being here is the same as being there (I'm an ASEL, commercial, instrument, glider, 3000+); but you gotta' admit guys, .....................it's getting BETTER and BETTER.Kingfish :( :(

I tried it and observed worse performance. This is on a i7940 oc to 4 ghz. I have seen good results with HT off and a setting of 14 but too close to call compared to no affinity mask at all.

I tried it and observed worse performance. This is on a i7940 oc to 4 ghz. I have seen good results with HT off and a setting of 14 but too close to call compared to no affinity mask at all.

Exactly the same as my observations. i7 975 4.5Ghz.Pete

I made some tests with HT on and off and different AffinityMask settings some time ago. I measured average fps in FSX with very high display settings over an urban dense area (London). After many hours of testing I came to the conclusion that enabling HT didn't improve performance in FSX. Different settings of AffinityMask didn't improve perfomance.I go along with Nick regarding the placebo effect. Over the years with FSX many posters have experienced increased performance in FSX with every new graphics driver, DX9 drivers, fsx.cfg tweak, OS tweak. I guess by now they would be able to run FSX with maxed out display settings!

I made some tests with HT on and off and different AffinityMask settings some time ago. I measured average fps in FSX with very high display settings over an urban dense area (London). After many hours of testing I came to the conclusion that enabling HT didn't improve performance in FSX. Different settings of AffinityMask didn't improve perfomance.I go along with Nick regarding the placebo effect. Over the years with FSX many posters have experienced increased performance in FSX with every new graphics driver, DX9 drivers, fsx.cfg tweak, OS tweak. I guess by now they would be able to run FSX with maxed out display settings!

Ulf,Call it what you want. It is not a placebo on my system. As I stated earlier in this thread, I see no increase in FPS on my machine. However, there is a definite improvement in the overall smoothness of FSX. I run locked at 30 FPS (I am constanly locked at 30 unless in a complex aircraft at a large airport). I have also found that with the i7 and a properly tuned system, FPS_Limiter is not needed and will actually induce stutters. Initially nobody believed me when I found that on certain systems increasing the BufferPools value to a very large number could help those with sound issues, autogen spikes, and stutters when running FSX on high autogen settings. Now many people including yourself use this tweak. I've got about 10 hours invested thus far trying various combinations of the AffiniityMask on my new i7 rig and thus far 85 with HT enabled is noticeably different in overall smoothness. Enabling hyperthreading on my machine required a voltage increase on my CPU to keep the system stable which means I am now running hotter under load but still safe. Maybe those that are finding a difference are not using the FPS_Limiter or are running Vistax64 or runnning on Win7 who knows what the difference is at this point.You said it before FSX is a weird bird. There are alot of variables between systems and configurations and if there is one thing we have learned about FSX, what works for one does not necessarily work for all.I am going to put Pete's suggestion of AffinityMask of 84 with HT enabled to the test and see what happens.

Ulf,Call it what you want. It is not a placebo on my system....You said it before FSX is a weird bird. There are alot of variables between systems and configurations and if there is one thing we have learned about FSX, what works for one does not necessarily work for all....

OK, that might be true, but it isn't my experience. :( I'm running an old i7 940 (revision C0/C1) and had a hard time to reach 4GHz. So to be able to do the HT and AffinityMask tests I had to decrease my OC before running the comparison tests. Maybe the effect of enabling HT and setting AffinityMask = 85 will only have a positive effect at very high clock speeds?Still IMO over the years with FSX I found that most peoples advice on tweaks and info on perceived performance gains after updating drivers have not been founded on facts. I do respect yours, Nicks and some other experts advice on tweaks and I do put them to a test on my system. But as I said: I go along with Nick regarding the placebo effect. Over the years with FSX many posters have experienced increased performance in FSX with every new graphics driver, DX9 drivers, fsx.cfg tweak, OS tweak. I guess by now they would be able to run FSX with maxed out display settings!I'm not an expert on hardware or FSX, but I do my benchmarking of FSX performance before and after every driver update and change of FSX configuration.I didn't follow the advice of some so-called "experts" when I upgraded my graphics card from a 8800GTX to a GTX285 on my present system. I got a 25% performance increase in FSX when flying over dense urban areas. I will probably upgrade my card in a couple of months to either a nVIDIA 300 or AMD Radeon 5xxx card. Maybe I'll thow in an i7 950 as well (and redo some HT/AffinityMask testing :-) ).

Over the years with FSX many posters have experienced increased performance in FSX with every new graphics driver, DX9 drivers, fsx.cfg tweak, OS tweak. I guess by now they would be able to run FSX with maxed out display settings!

Lol...so true. :(

Exactly the same as my observations. i7 975 4.5Ghz.Pete

Brings no improvement here, either. i7 975 4.4GHz, Win 7 RTM.Tim

Brings no improvement here, either. i7 975 4.4GHz, Win 7 RTM.

I think that's pretty much what we meant by the phrase "too close to call". ;-)Pete

Please let me make something perfectly clearPlacebo effect is real and it can be applied to any test in which a human is judging or gauging a result. It will always be part of a test result when dealing with something like thisWhen I see people swear by perf changes in things like Fiber Frame Time Fraction on systems running multicore processors and FSX SP1/2 and other fsx.cfg edits/tweaks that have no bearing on a result I must question the rest of the suggestions being made. FFTF has no impact with multicore processors and therefore what ever the person is seeing has nothing to do with that edit. Suggesting others do it and seeing 5 come back and post they also see an improvement is placebo. However I never said working with Affinity Mask will provide nothing but a placebo effectWindowsXp does make use of HyperTransport with a patch but not as much as Vista/W7What I have said is the HyperTransport will not in any way increase FSX performance by processing its application threads through the virtual caches and that is a hard fact. I have also stated that it is possible to smooth FSX performance by tuning out thread collisions and in some cases by removing the app from the primary core and that IF such stutter problems exist on a system, the process to successfully remove said stutters can be different from system to system and that has been true since the Intel C2/Quad was released.I highly suspect there is a common denominator to systems that find i7 HT ON and AM=85 useful. What that is at this point is an unknown but I do suspect it is centered on a VISTA/W7 OS process or driver managementWhat happens with these "I found the cure" threads is there is a defined number of people who actually DO see an improvement, another defined set that see WORSE performance, and yet another set that see nothing. Then there is always a number of people who say they 'think' they see a difference and those are usually the placebo's. If you can