Like many of you, FSX runs bad on my pc with low settings. I have an amd 64 3200, 2gig ram and a 6800GT(256mb).My question is, what hardware upgrade would help alot in my quest to run FSX in medium detail?I'm planning to upgrade my cpu to an amd 4600X2 dual core. Would that help my fps, or would it just be a waiste of money? The rest of my configuration would remain the same.Is cpu more important then a gfx card (for FSX)?

Take a look here: http://www.forum.aerosoft.com/viewtopic.php?t=8707It's a opinion of course but a very intersting and educating one with respect to hardware requirements for FSX. My opinion. I'll stichk to FS9 for awhile until DX10 is on the market and hardware gets cheaper.Rob "Holland&Holland" de Vries http://fool.exler.ru/sm/fly2.gif"To go up, pull the stick back. To go down, pull the stick back harder"

Good link Rob. A very level-headed set of recommended system components.Gary

I'd say you'd see a 25-30% boost in FPS in FSX, plus a lot smoother overall performance with the second core handling most non-FS related processing, with your proposed CPU upgrade. It's your call as to whether this is worth it for you.From what I have seen with FSX, GPU plays a much greater role in FSX than with previous titles, but CPU still remains the main determinant of overall FSX performance.Gary

Yes, I have to say that the CPU upgrade would help some. A faster video card would as well (I assume you're on AGP though, so that kinda limits you.)From what I've seen, the 6000 series cards aren't handling FSX as well as the 7000 series. The 8800 models (DX10) are due out in a few weeks so that should help the price on the 7000 parts to come down a bit.Finally, RAM. You have 2 gigs already. If you still have 2 slots open, get 2 more. Flight sim needs memory to put all those autogen objects in. I consider 2GB the MINIMUM for FS now if you want to have autogen on.To give you some perspective (on the FP Beta) I am getting a pretty decent 15-20 FPS on 1280x1024 with all the scenery (except water which is at 2.x low) all the way to the right, autogen on Dense, and the AI traffic set to medium while flying around KATL. This is without doing any of the various tweaks posted around the forums here, just strictly messing with sliders. I'm attaching 3 CFG files that I put together last night, each with increasing scenery + autogen (and obviously with decreasing framerate.) All were certainly flyable.My System:Ath64X2 4800+ AM24 GB DDR2-800 Corsair Memory (4-4-4-12-2T)2 XFX GF7900GT's (stock clocks) SLI'ed.4 Seagate 7200.9 300GB SATA2 drives in RAID 5Asus M2N32SLI-Deluxe MB158349 gets me a steady 20 FPS all around KATL and downtown.158350 gets me 17-20 all around KATA + downtown158351 gets me 12-20 in the same areas and 20 when I'm away from the city (I tried it out around Daytona Beach and it locked on 20).

>Finally, RAM. You have 2 gigs already. If you still have 2>slots open, get 2 more. Flight sim needs memory to put all>those autogen objects in. I consider 2GB the MINIMUM for FS>now if you want to have autogen on. Are you sure about that claim -- that FS really wants more than 1 gig, in particular for autogen? Have you tried it out? I've got 1 gig (DDR-400 dual channel); and a monitor shows that while playing FSX, the ram is definitely used up to 100% but I was really wondering if getting some more ram would be useful. I can go to 2GB for $120, but would rather not do this if it's not going to make a giant difference.I'm not considering cpu upgrade (amd x2 3800) nor video upgrade (x850 PE AGP) btw.

In my experience, FS9 is happy to wallow around in 1G memory, even with a bunch of addons running in the background. FSX, on the other hand, has shown to be markedly smoother with 2G vs 1G RAM, and that's as a stock install. When addons enter the scene, 2G will be a certainty. I don't foresee any tangible benefit to either sim with 4G RAM, even with addons.Gary

not to cause a fight here, but--It does not appear that FSX uses more than 2GB of memory in the testing I have conducted. In a single monitor situation, accepting all defaults and without using any tweaks, I have found that memory "pages" only occasionally and page file utilization is virtually nil. I have also found that disc access is also random and sporadic-- it is not intensive enough to be able to support with data the concept that a faster drive will give better performance in flight (it may decrease initial load times). Processor utilization will be 100% on "cpu 0" and will spike occasionally on "cpu 1" in hyperthreading CPUs, but will generally only use a very small percentage of the second cpu core or hyperthread.At least this is what I have found with testing...Data is attached. methodology for gathering data was to initiate a flight in fair weather using the "ultralight" and set the ultralight into a slow, banking turn. Once a stable turn was established, I started PERFMON to log data every 5 seconds for 5 minutes. I stayed "hands off" of FSX. After data was gathered, I rebooted the machine and restarted the same flight and profile. Again started PERFMON with the same data collection parameters.Data attached is from a P4 3.3Ghz, 800 mhz fsb, 2 gb ram. 256 mg NVIDA 6200. 3 Hard disk configuration with WIndows on IDE 0 Page File on IDE 1 and FSX running on a SATA drive. Machine was defraged prior to start of testing.File is CSV format.Does anyone else have similar data?