Hi,I need your help in understanding the poor FSX performance on one of my computers. I have 2 similar computers running the exact same OS (Win 7 Home Premium 64-bit) with 4 GB DDR2-800 RAM , same Softwares and both have FSX Acceleration (including exact same addons) on a dedicated SATA hard drive (WD 7200 RPM 320GB and WD 7200 RPM 160 GB). No tweek applied to FSX.CFGI built both systems myself, following the same steps and sequence when installing OS, drivers and softwares. All drives show 0% fragmentation on both system (I defrag regularly using Win7 built-in defragger). BIOS settings are identical.The only major difference between the 2 systems is the CPU and video card. System 1: Asus P5Q (P45 chipset), C2D Quad Q9650 3.0GHz with Nvidia 8800GTX 768 MB drivers 182.50 (Vista 64) 1280x1024x32 (TH2G digital using single display mode only during comparison).System 2: Asus P5Q SE (P45 chipset), C2D E8400 3.0GHz with Nvidia GTX250 512MB drivers 182.50 (Vista 64) display 1680x1050x32.Previously, under Windows XP SP3, FSX performed almost the same way on both computers. But since moving both to Windows 7, there is a big difference in FPS and smoothness between the two using the exact same settings (FSX and drivers) and the exact same flight situation.My flight situation puts me on the threshold of a medium size airport in the default BE58. System 2 is great, with average 45 FPS, no fluctuation, no pause and no stutter.System 1 is the one giving me problem, with FPS fluctuating between 6 and 25, with lots of micro pauses and stutters. FPS really fluctuates a lot.Let me mention again that I am comparing using the same settings and flight situation on both computers.On system 1, I noticed that the "System Idle Process" always uses between 50 and 85% of CPU cycle when FSX is running, leaving FSX with 15-50% (mostly 20-25% average). System2 always shows FSX using 85-90% of CPU cycles.I believe that FSX could perform better on system 1 and probably without pause or stutter if FSX was using more CPU cycles. Out of curiosity, I replaced the video card on system 1 with a ATI HD5850 1GB DDR5 card but the performance remains the same (Cleaned the NVidia stuff with driver sweeper). What could be the reason for my system #1 to give me worst performance (almost 50% less) than the slightly less powerful system 2 ?I appreciate any thought you may have.

On system 1, I noticed that the "System Idle Process" always uses between 50 and 85% of CPU cycle when FSX is running, leaving FSX with 15-50% (mostly 20-25% average). System2 always shows FSX using 85-90% of CPU cycles.I believe that FSX could perform better on system 1 and probably without pause or stutter if FSX was using more CPU cycles.

Hi, for clarification System Idle process is the time that the CPU is idle so 100 is better than 0 (to speak in your terms). I am confused by your statement, if idle was 50-85 on system 1 then you would leave by your analogy FSX with 85-50 or you could say FSX is using 50-15.You say system 2 shows FSX using 85-90 not sure what you mean here. If you are referring to system idle again then system 2 is leaving FSX 85-90 or you could say FSX is using 15-10.Therefore system 1 indicated that FSX used more than system 2 at least that one time but both are pretty close if it was a momentary spike in cpu usage. Certainly not enough to detract that one system uses more CPU than the other by that test.It could be a host of issues from how the MB handles a quad versus a dual CPU to a driver install issue (even though the same maybe something did not install right) weakness of hard parts (something wearing out/not working right).I often wondered about FSX and SP1 and SP2 and how they truly interacted when spreading CPU load over multiple cores or increased use of virtual address space. The FSX program was not originally designed to do that and I am not sure you can simply change the address aware flag in a service pack or simply code the program to use multiple cores in a service pack and get a good result. The service packs were obviously after thoughts that were attempting to correct major performance issues as opposed to FSX being designed correctly to begin with.As to why the differences...in your test? Too many variables to mention. It could as I say be a host of anything

I am confused by your statement, if idle was 50-85 on system 1 then you would leave by your analogy FSX with 85-50 or you could say FSX is using 50-15.You say system 2 shows FSX using 85-90 not sure what you mean here. If you are referring to system idle again then system 2 is leaving FSX 85-90 or you could say FSX is using 15-10.Therefore system 1 indicated that FSX used more than system 2 at least that one time but both are pretty close if it was a momentary spike in cpu usage. Certainly not enough to detract that one system uses more CPU than the other by that test.

Thank you Gary for your reply,I was referring to the CPU usage as displayed in in the "Processes" tab in Windows Task Manager. When, during my test, I saw that FSX on system 1 (Quad core) was not smooth and had low FPS compared to system 2 (C2D), I openned Task manager thinking I could identify what process was preventing FSX from running better. I saw the FSX was showing between 15% to 50% CPU usage and the only other process showing a CPU usage was "System Ile Process", showing a 50% to 85% of CPU usage.I would not have worried about it if FSX was running smooth with FSP about the same (I would understand a slight difference between system 1 and 2, but not half the FPS).

It could be a host of issues from how the MB handles a quad versus a dual CPU to a driver install issue (even though the same maybe something did not install right) weakness of hard parts (something wearing out/not working right).

That's true. I am just surpised by the result considering that when both systems were running XP Pro SP3, I was having similar performance and no stutter or pause with both systems, system 1 (Quad core) being slightly better with texture load at the time. I would imagine that if FSX uses only 50% of CPU cycle on the Quad and that "System Idle Process" is the only one using the rest, then that's all FSX needs to do its tasks. But why do it with such a low performance compared to the C2D. I guess that is what really puzzles me.I know that my test is not a scientific one, but I took great care in setting up both system the same way (just as I once did under XP Pro SP3).If a friend of mine were to come to me saying he has "system 1" and claimed that it performs less than my "system 2", I would probably tell him "sorry body, no way you set it up the same way as mine, otherwise you would get similar performance. What mistake did you make ?". See the irony ?Anyway, thank's for your input. I am in no way claiming that a lessor system outperforms a better system. All variables being the same (clock speed, RAM, etc..) a Quad should perform at least equal to a C2D with FSX, if not slighttly better with texture load (That was my observation under XP Pro). FSX on system 2 (C2D) performs as well in Windows 7 as it did in XP Pro. I am simply trying to achieve the same on system 1 (Quad).Take care !"

I found the solution to my problem in the tread "The Mystery of frames dropping in w7"http://forums1.avsim.net/index.php?showtopic=275281Solution: Windows 7 Control panel/Power Options: set it to High Performance.System 1 (Quad) power option was set to "Balanced", while system 2 (C2D) was set to "High Peformance". I was quite sceptical that it could have an impact on how FSX runs, but I had to try. I changed system 1 to "High Performance" and VOIL

I found the solution to my problem in the tread "The Mystery of frames dropping in w7"http://forums1.avsim.net/index.php?showtopic=275281Solution: Windows 7 Control panel/Power Options: set it to High Performance.System 1 (Quad) power option was set to "Balanced", while system 2 (C2D) was set to "High Peformance". I was quite sceptical that it could have an impact on how FSX runs, but I had to try. I changed system 1 to "High Performance" and VOIL