>OK, I will come back to this... you are still trying to place>a linear value on FSX performance using my system as the>target goal..How so? The argument I posed has to do with some practicallity issues. As I had said before, I don't like overly cluttered autogen, and I'm not particularly interested in AI traffic. I am very interested in clarity, and detail in the terrain textures, sans objects per se. I think there is clearly an answer to this question, I just don't know what it is yet. You have the capacity to test this. What say you?Wipe out ALL autogen & traffic, and clouds. How to texture tiles look now, with best testbeds? I think it's a very valid question. I can try this myself, but need to install Vista 64 first, which is what I plan to use for FSX, though I guess I could dabble now in XP. I have an unopened copy of V64 . . .

Noel and Nick,Re:>If YOUR particular config of FSX is say 100%, then at what>LOD, slider settings, autogen, etc, relative to that would>allow my cluncky memory sub to display the same visual>CLARITY, offset by reducing visual COMPLEXITY? I think this>is a valid question. Certainly the demand on the memory sub>is affected by how we set up FSX. If we could assign a value>to the capacity of your rig to process the load, what value>would we assign my system in terms of CPU/Memory sub? I'm>back to values, but surely this IS a quantity that could be>measured. Put another way, there should be a point at which>WE can BOTH dial back complexity, and the displayed experience>will be the same. If we put all sliders LEFT, 1024 x 768, our>systems should be displaying the same result, or am I wrong in>this assumption? If this is not the case, then there are>issues I def do not comprehend. Where is the point at which>the DDR3 system, same CPU, etc, begins to depart from the>optimized DDR2 system?I think this is a valid, if not THE, question with regard to this whole tweaking game. Certainly, Nick's test results show that DDR3 did not trigger FSX's load shedding routines, whereas the DDR2 did. So at what FSX settings are DDR2 and DDR3 able to produce the same visual output? Is it just the slight notching down of one key FSX setting, such as autogen, enough to free the FSX engine from overload and bring this equality of visual output or are there a few notable key setting compromises that need to be made? If the former, then perhaps DDR2 is a suitable price/performance compromise for most of us given that it produces almost the same visual output. Just another way of looking at this complex issue :-)Gary

Certainly, Nick's test results show>that DDR3 did not trigger FSX's load shedding routines,>whereas the DDR2 did. So at what FSX settings are DDR2 and>DDR3 able to produce the same visual output? Gary, Are FSX's load shedding routines configurable directly, or is it more of a case of just adjusting detail, traffic, autogen, then the game engine assumes complete control?

Nobody really knows for sure Noel. The only thing we have to go on is the occasional hint from MS staff that load shedding is definately happening once 10 FPS or lower is hit, but at what point this load shedding cuts in for each setting is anybody's guess.Gary

hint from MS staff that load shedding>is definately happening once 10 FPS or lower is hit, but at>what point this load shedding cuts in for each setting is>anybody's guess.Hmm, that is signficant, kinda sorta . . .QX9650 w/ Retail HSF|8GB Muskin PC-6400|ASUS P5E|EVGA 8800GT @700|Seagate SATA 2 x 4|Seagate Cheetah 15K.x|XP Pro|Vista 64--soon to be installed

This was exactly my point to Noel about trying to measure and adjust based on some type of mathematical equation is quite impossible, for the most part because we do not have enough information about how the application is designed. I also think they made some shifts in where load shedding occurs between RTM, SP1 and SP2. Those who are on RTM typically set up one way, SP1 another and SP2 yet another routine to trim in the sim. When I ran the benchmark I noticed road traffic got a low priority however if I fly a different route just a mile or so to the east and get out of the load zone the road traffic volume returned. The entire time frames remained well above 10 however during the flight a mile from the benchmark flight path, frames were an average of 10 higher.

There is a huge difference between system tuning and FSX tuning. One is not related to the other however we don

Many thanks for those setup ideas. I installed FSX, SP1, GEXn. Haven't downloaded the other two recs. Here is a very compressed pic flying over SF. I looked closely at texture quality and things look very good. I don't think you can really see this in these compressed pics, but I can attest it's good. Any way I can paste a native, uncompressed image here? This is full everything except for vehicles n traffic which are about 15%, and unlimited frames, just to see what it could do. As with FS9, I will dial back autogen and frame lock for obvious reasons. Anyway, I am shocked at how good this both looks and flies. It is VERY smooth, silky. Of course this is with a lightweight aircraft, but so far that's what I'll use this for, VFR and lighter aircraft for VFR etc. Anyway, yes I'm sure better memory will help, but no, I don't believe this is critical. Everything looks totally fine. I am running at 4.1GHz, with my old memory. http://forums.avsim.net/user_files/187641.jpgLittle pic showing texture quality on the golf course:http://forums.avsim.net/user_files/187642.jpgIs there a correct order for installing the other addons?

>Anyway, yes I'm sure better memory will help, but no, I don't believe this is critical. Glad you feel vindicated in not spending money. That

>Glad you feel vindicated in not spending money. That

Sorry I was a bit short this morning Noel. My daughter and grandson were involved in a collision this morning and although they are alright, I was really in no position or mood to assess a post.When you make comments like what I quoted above they have a conclusion ring to them unless qualified, which your post did not, nor did it say you needed to do more tests. The way it was framed you were drawing conclusions. You may have also been in the process of adding/editing when I stopped by this morning.. none the less my initial read of it left me thinking you had drawn a conclusion.Understood on the checks and tests, and yes its fine to install UTX after GEXn however you must rerun the GEXn/UTX texture selection again and make sure to defrag after a huge install like GEX/UTX

Hey Nick,I saw your pics and I think they speak for themselves. You can see my current rig below. I just ordered http://www.newegg.com/Product/Product.aspx...N82E16813131200with http://www.newegg.com/Product/Product.aspx...N82E16820227292and I am going to move the E8500 over with everything else. Basically a mobo/memory only upgrade. I figured with this processed I might as well just go for it.Thanks for the motivation, (PS there is an Asus P5N32SLI-Plus/E6600/2 gigs Corsair XMS2 for sale cheap LOL)-PaulPrimary RigLiquid CooledIntel C2D E8500@ 3.8Asus P5N32SLI-Plus2 gigs Corsair XMS PC6400 4 4 4 12 @936Dual OC'd XFX 8800GTX @ 2 gigs24 inch Widescreen LCD 16XAA/16XAFDual 19 inch LCD'sRaid-0+1PCPower and Cooling 1k Quad SLIhttp://home.comcast.net/~psolk/3monitorsa.htmlBackup RigAMD 4000 San Diego @ 2.72 Gigs Kingston Corsair XMS CL2XFX 7900 GTX Raid-0

Doing some observations using memory scores of 54.3ns/9300mb/s (OCZ DDR2-1000 @ 1056) vs DDR2-800 @ 808 with scores of 59.6ns/8350mb/s, No practical differences are observed. Of course, this is unquantified. But I can say perf in FSX at 4.1Ghz really is identical, without spliting hairs by doing more detailed & controlled comparison. I am running most sliders full right, 1600x1200. I am also flying Cessna 441 Conquest in FSX, very high autogen density, flying around the SF Bay Area. Performance is very smooth, terrain detail (1m mesh/7cm texture resolution) seems excellent. I can see textures updating quite a ways away. Frames from takeoff at KSFO, circling around the bay and over downtown, are around 26 to 38 with very dense autogen, terrain distance to max. With default slower aircraft, 32 - 45 or higher.I am using SP1 only with FSX. Should I install SP2? Is it uninstallable? I am using XP, DX9.I will do some more stressful tests to try to tease out any practical differences. I'm guessing the differenses in the memory perf scores using Everest is just to small to translate to realworld, real signficant differences. It's hopeful for me tho as I don't see swapping out mainboard and memory as perf is certainly . . .better than expected. If I find someone without FS needs who needs an upgrade who doesn't need the highest end bandwidth, I'll change then. But at least I'm finding this system as it is currently, does quite well. I think what I have discovered is what I had suspected: dialing down a slider, a notch or two, WILL translate to setting up the sim to work well to meet my particular perf criteria, which includes excellent visual clarity, detail, and smoothness, though I haven't had to downturn sliders very much so far.Nick, what are your latency & read scores? Yes, I know, they are what they are, benchmarks. Even so, they should represent some scalability applicable to FSX, even if the way to measure enhanced performance is not so easily measured. Visual clarity, smoothness, frames, and complexity are hard to quantify . . .Noel

Nick, question if you will . . .I have an op to unload my DDR2-800 4-4-4-10 (rated 5-5-5-18) modules for DDR-1066 5-5-5-18 from mushkin. I am running my current modules at 400-404 FSB, 1:1 ratio, tRD of 7 is the best it goes, with latency of 59.5ns, read rates of 8350mb/s. Can you take a stab at guesstimated scores with the 1066? It will cost me about another $100 for the 8gb I would hope to install again. From the sound of it, I am fortunate to have 8gb run stable and with modules performing perhaps better than they might be from other batches of the same product. So I guess it's possible I could get another 8gb that do not run stable together, I don't know.Anyway, venture a projection? I imagine the timings in the 1066 will not do 4-4-4-10 of the DDR2-800 modules I have now. Could be a wash do you think?QX9650 w/ Retail HSF|8GB Muskin PC-6400|ASUS P5E|EVGA 8800GT @700|Seagate SATA 2 x 4|Seagate Cheetah 15K.x|XP Pro|Vista 64--soon to be installed

I really do not know about those sticks and what they can and cant do Noel I would need to do in-depth research myself and probably contact the memory company too. I do spend some time doing that when I purchase but not hours on end. That is why I pay the premium prices and do not screw around with the cheaper modules. I purchase them overrated on purpose. Many times its the luck of the draw and the bin the chips come from that make the grade. Many times you can contact a rep and discuss what you are asking with them because they usually know which bin runs are hitting the numbers and which ones are not. And they are going to tell you the same thing I will... running that much memory is usually the luck of the draw to hit specs even on the better bin runs.I am very busy with GEXn Europe right now so I have not had much time to forum hop. Getting back to what you asked 2 posts up,.. forget it, I have said it over and over again.. you are not going to find a scale that will give you a curve... This forum system deletes the attachments after the thread goes 2 pages which sucks because my presentation is now gone with my readouts. My 'industry standard' benchmark values mean squat unless you know how to read them. Here is the image of the last run I made with that towerhttp://forums.avsim.net/user_files/187728.jpgOK, so I can have 12000 read values @ DDR2 1066 and a latency of 50ns and still not be as fast as DDR3 1440 running 1T @ CAS6. What I have been trying to convey is that this is the ENTIRE ball of wax that produces the result.. the CALWI is #1.. get that as low as possible at the highest memspeed. The tRD is the next, get that as low as possible at the highest memspeed and lowest CALWI, 1T is next.. if that