HyperThread - FSX - i7 -and YOU

[Guest Seria17hri11er]

FSX does not make any use of hyperthread - Phil TaylorFiber frame time fraction is useless with multicore processors after SP1 - Phil TaylorAffinity mask can be used to tune out thread collisions and most will find its not needed - NickN

Good to know. It appears that his settings are giving him FSX performance gains indirectly, no?

[Guest Nick_N]

Conditioning and Distraction: Why we see and believe what we want to In Simon's completed experiment, which involved two videos ( the alternate with a woman holding an open umbrella over her head walking amidst the two teams of players ) and differing task levels, gradating from easy to hard, only 54% percent of the observers noticed the unexpected event and a remarkable 46% never saw a thing while attending their observational task.Observational Blindness - http://www.wjh.harvard.edu/~cfc/Simons1999.pdf With MSFS the operand in motion is the user reading something and making a settings change. MSFS users are litterally desperate for better performance. One day they post everything is perfect, the next they post a change has made it better, and then a month later, another change has again made it better. They then assume automatically a settings or config edit will: a. Make a positive differenceb. Make a negative differenceThe third is: Make no difference.. and because of the high desire to want to see something happen, the user will then fall in the 54/46% catagory of Observational Blindness usually to A and Band if we are talking about a group, most users can be wrong about what they think they see with only a small portion assessing the correct screen result. The items in question do nothing for FSX, at all. The only thing that may have changed is thread collisions due to changes made in the affinity mask setting in combination with tinkering with HT, past that there is nothing in that system being upgraded by HT or Fiber Frame Time Fraction and nothing worth setting a system up like that to evaluate. Anything 'truly' different in a positive outcome can be accomplished through correct tuning means. Fiber frame has NO, none, zip impact on multicore since SP1 as Adam from Aces clearly explained:----------------------------------------------------------On multi-core machines in SP1, we moved many fiber jobs off of the primary thread and onto secondary threads. Since FIBER_FRAME_TIME_FRACTION only affects scheduling of jobs on the primary thread, it will have less of an impact on the performance of Flight Sim on multi-core machines. In fact, we moved so many jobs off of the primary thread that there probably isn

[AlphaInfinity 13]

Conditioning and Distraction: Why we see and believe what we want to In Simon's completed experiment, which involved two videos ( the alternate with a woman holding an open umbrella over her head walking amidst the two teams of players ) and differing task levels, gradating from easy to hard, only 54% percent of the observers noticed the unexpected event and a remarkable 46% never saw a thing while attending their observational task.Observational Blindness - http://www.wjh.harvard.edu/~cfc/Simons1999.pdf With MSFS the operand in motion is the user reading something and making a settings change. MSFS users are litterally desperate for better performance. One day they post everything is perfect, the next they post a change has made it better, and then a month later, another change has again made it better. They then assume automatically a settings or config edit will: a. Make a positive differenceb. Make a negative differenceThe third is: Make no difference.. and because of the high desire to want to see something happen, the user will then fall in the 54/46% catagory of Observational Blindness usually to A and Band if we are talking about a group, most users can be wrong about what they think they see with only a small portion assessing the correct screen result. The items in question do nothing for FSX, at all. The only thing that may have changed is thread collisions due to changes made in the affinity mask setting in combination with tinkering with HT, past that there is nothing in that system being upgraded by HT or Fiber Frame Time Fraction and nothing worth setting a system up like that to evaluate. Anything 'truly' different in a positive outcome can be accomplished through correct tuning means. Fiber frame has NO, none, zip impact on multicore since SP1 as Adam from Aces clearly explained:----------------------------------------------------------On multi-core machines in SP1, we moved many fiber jobs off of the primary thread and onto secondary threads. Since FIBER_FRAME_TIME_FRACTION only affects scheduling of jobs on the primary thread, it will have less of an impact on the performance of Flight Sim on multi-core machines. In fact, we moved so many jobs off of the primary thread that there probably isn’t enough fiber work left to soak up the full time allowed by the default value of 0.33. Therefore, on multi-core machines, there’s very little reason to tweak the fraction because it really only impacts performance of single core machines.Please emphasize the pointlessness of tweaking this value on multi-core machines. Further, setting it to values greater than the default value on single-core machines can increase frame rate volatility because it increases the amount of time we *might* allocate to fibers if there is adequate work waiting in the queue. When the queue is full, we’ll allocate the full amount of time to fibers but when the queue is empty, fibers get no time because there is no work to do. If the rate of new jobs entering the fiber queue is bursty and the full time allowed for fibers is large, then you can imagine how this would increase volatility. If people feel like the fibers aren’t getting adequate resources, they would be better off leaving the fiber fraction alone and just lowering the frame rate limit slider, which helps divert more CPU time to primary thread fibers without increasing volatility.=======================================Therefore, a user who does not have any concept in technical aspect applied to what they doing to the system with edits/tweaks will in turn display a high probability of completely inaccurate assessment in the reported result with respect to those changes.Bob's assessment of thread collisions and the primary core was a good one, because he has a high level of education and experience in the electronics field and has the background to look at the readouts/result and draw a conclusion which comes from a solid basis in fact, not fiction. The moment we want to believe something, we suddenly see all the arguments for it, and become blind to the arguments against it. -George Bernard Shaw

Nick,If you had to rate the top 3 changes to the config file in FSX in terms of overall performance impact(outside of the usual slider changes in FSX), what would they be?

[Guest Nick_N]

There are only 4 changes in that config file that are going to make any difference and they are all problematical to the system running the application with the exception of Disable PreloadTexture Bandwidth - usually 80 is the sweet spot but can be anywhere from 40 to 80Disable preload - always = 1Bufferpools - defined by the system, either no entry or a value between 35000000 and 200000000. Anything above 200MB with the 1GB video cards that displays a positive result indicates a latency issue on the buss the higher value assists in reducing which is usually clued to the user when sound crackle appears with high autogen sliderAffinity Mask - most dont need this entry at all but there are exceptionsThere is no mystery hidden secret in that config file. Thats it, what you see above are the only edits that may make a difference and all of them are system dependentFiber Frame Time Fraction - is only used with single core procs and that edit is really useless after Sp1. As Adam from Aces pointed out, drop the frame lock.

[Guest Nick_N]

..and one other itemTerrain Autogen Trees and building restriction lines - I have never, ever used these on any modern system as they tend to make tuning a nightmare. The only systems I can see making any use of those lines are older AMD and Pentiums.

[Bob Scott 7,721]

I should point out here that the perf increases I observed when initially experimenting with HT enabled and various AffinityMask settings turned out to be owing to an unrelated and undiagnosed issue I was having with Ultimate Traffic II, which was loading at sim startup, and then surreptitiously closing (and taking all the traffic with it) 15-30 sec after the sim started. So the benchmarks I was comparing my "with HT" performance against had AI traffic, and the perf I measured with HT enabled and AM at various nonstandard settings was being inflated by the absence of AI traffic due to the then-undiagnosed traffic problem.Once the traffic situation was sorted out, and the tests re-run, I concluded that there's no appreciable difference between HT-enabled performance and HT-disabled performance at the same clock speed and with the same number of cores in operation. But given that I had to drop my overclock a notch with HT enabled, I concluded that HT actually decreases my performance after all is said and done.TANSTAAFL...RegardsBob ScottColonel, USAF (ret)ATP IMEL Gulfstream II-III-IV-VColorado Springs, CO

[Guest Nick_N]

I should point out here that the perf increases I observed when initially experimenting with HT enabled and various AffinityMask settings turned out to be owing to an unrelated and undiagnosed issue I was having with Ultimate Traffic II, which was loading at sim startup, and then surreptitiously closing (and taking all the traffic with it) 15-30 sec after the sim started. So the benchmarks I was comparing my "with HT" performance against had AI traffic, and the perf I measured with HT enabled and AM at various nonstandard settings was being inflated by the absence of AI traffic due to the then-undiagnosed traffic problem.Once the traffic situation was sorted out, and the tests re-run, I concluded that there's no appreciable difference between HT-enabled performance and HT-disabled performance at the same clock speed and with the same number of cores in operation. But given that I had to drop my overclock a notch with HT enabled, I concluded that HT actually decreases my performance after all is said and done.TANSTAAFL...RegardsBob ScottColonel, USAF (ret)ATP IMEL Gulfstream II-III-IV-VColorado Springs, CO

There is no such thing as a free lunch - Phil TaylorWhich is the exact conclusion I came to after several weeks of working with i7 and FSX earlier this year BobI think what these people are seeing in a 'smoother' result with HT enabled and manipulating AM is a combination of installed scenery and settings they are running and I do think the Affinity mask adjustments suggested may also be centered around Vista/W7 use of HT too. That is not to say it wont work to smooth things out and its not to say that what they are doing is wrong.. however I do think there are other ways to trim out what they are seeing and not lose clock speed to HT, a feature that adds nothing to the simNow, I will say this.. Windows7 does make full use of modern HT over Windows versions of the past however that will not translate into FSX use of the feature, only OS processes which I do not see as being + or - for FSX since a correctly set up system is not overrun with them. FSX was not made for Windows Vista or W7. It does not comply to Vista memory and driver management standards and although it is 'certified' to run on Vista by MS it is not 'optimized' for it.. FSX was made under Windows XP driver and memory management standards... which is why I use XP x64 and see no reason to change that at least for now.Thanks for reporting your findings sir! Its always a pleasure and very refreshing to read your assessments in evaluations like this. :(

[Guest Seria17hri11er]

There is no such thing as a free lunch - Phil TaylorWhich is the exact conclusion I came to after several weeks of working with i7 and FSX earlier this year BobI think what these people are seeing in a 'smoother' result with HT enabled and manipulating AM is a combination of installed scenery and settings they are running and I do think the Affinity mask adjustments suggested may also be centered around Vista/W7 use of HT too. That is not to say it wont work to smooth things out and its not to say that what they are doing is wrong.. however I do think there are other ways to trim out what they are seeing and not lose clock speed to HT, a feature that adds nothing to the simNow, I will say this.. Windows7 does make full use of modern HT over Windows versions of the past however that will not translate into FSX use of the feature, only OS processes which I do not see as being + or - for FSX since a correctly set up system is not overrun with them. FSX was not made for Windows Vista or W7. It does not comply to Vista memory and driver management standards and although it is 'certified' to run on Vista by MS it is not 'optimized' for it.. FSX was made under Windows XP driver and memory management standards... which is why I use XP x64 and see no reason to change that at least for now.Thanks for reporting your findings sir! Its always a pleasure and very refreshing to read your assessments in evaluations like this. :(

XP 64 bit is better than 32 bit for FSX because it supports a higher amount of RAM right?

[Guest Nick_N]

XP 64 bit is better than 32 bit for FSX because it supports a higher amount of RAM right?

Yes, and it supports the extended address space needed for FSX SP2, 4GB+ physical memory and the large memory video cards

[fsape 0]

There are only 4 changes in that config file that are going to make any difference and they are all problematical to the system running the application with the exception of Disable Preload

Hey, Nick, just a quick word of thanks from me. I've been following these forums for years, and your tips and suggestions on tuning FS and general tech insight into FS, has been truly useful and valuable. So, thank you!

[Guest Seria17hri11er]

Hey, Nick, just a quick word of thanks from me. I've been following these forums for years, and your tips and suggestions on tuning FS and general tech insight into FS, has been truly useful and valuable. So, thank you!

No doubt, ditto!

Yes, and it supports the extended address space needed for FSX SP2, 4GB+ physical memory and the large memory video cards

So, about "large memory video cards." What exactly do you define by large? I ordered a MSI OCv3 GTX 260. 896 MB of RAM. Running this card with a 32 bit OS will result in the system not completely utilizing all the RAM? I intend to upgrade to W7 64 Bit when it comes out this month.

[Guest AVN8tr]

Well since it always seems I'm disagreeing, I'll add that HT is absolutely useless... but you don't need to turn it off these days unless of course you still have a P4. I always liked this saying "Threading is not something you do to impress your neighbors"... Also... "threading is not for performance, its a method of providing responsiveness to an application". never confuse threading with multiprocessor support, it'll only get you in trouble.

[monsterdonster 0]

Hello,sorry for digging this thread out, but I have some serious issues with my FSX. It stutters heavily even running the default flight with the Trike at Friday Harbor. The Fps jump permanently between 4 and 1xx. Especially while panning fps drop into nearly nothing while some strange graphic errors occur. Looks like the autogen drops from the ceiling. My FSX is setup like told by Nicks guide. The strange thing is, sometimes FSX runs absolutely smooth as if the problem has never been there. The next day it stutters all over again.My SystemIntel Core i7 920@ 3,4 GhzGeforce GTX 275 2GB6 GB RAMWindows 7Help would be appreciated, as I

[Sesquashtoo 2,846]

Hello,sorry for digging this thread out, but I have some serious issues with my FSX. It stutters heavily even running the default flight with the Trike at Friday Harbor. The Fps jump permanently between 4 and 1xx. Especially while panning fps drop into nearly nothing while some strange graphic errors occur. Looks like the autogen drops from the ceiling. My FSX is setup like told by Nicks guide. The strange thing is, sometimes FSX runs absolutely smooth as if the problem has never been there. The next day it stutters all over again.My SystemIntel Core i7 920@ 3,4 GhzGeforce GTX 275 2GB6 GB RAMWindows 7Help would be appreciated, as I

[Guest WarpAir]

defrag might help.or maybe even uninstall and reinstall of service packs