[BufferPools] PoolSize=0 the holy grail of FSX performance...

[Kosta 122]

Well, I tried that setting in Nhancer. Didn't see any noticable difference.But, after I updated the driver to the last beta (see HW forum post), turned off C1E in Bios, left Nhancer on high quality, I get no more tree spiking!! Get a load of that...

[FalconAF 213]

so how would that translate into nhancer? What settings should I use?

And there's the rub, guys. One person says they aren't using nHancer...another one uses it. Lets' clear somethng up here, eh?If you use nHancer and follow Nick's settings, you are OVER-RIDING ANY settings made in FSX OR the nVidia Control Panel for the same-type settings. That's what the "Allow Override of Predefined Settings" option does in nHancer.So "what you do" or "what you don't do" will depend on:1. Am I even using an nVidia GPU to begin with? (If you are reading this and using an ATI card, I'd just quit reading it).2. Are you trying to use ONLY the settings in FSX for things like AA? (You don't use an nVidia Control Panel Profile for FSX? You DISABLED it altogether, I assume?)3. Are you TRYING TO USE a combination of "native slider or drop-down menu" settings in FSX, AND the nVidia Control Panel for an FSX profile, for the same things (AA, etc)? If so...WHY?4. Are you using nHancer? Do you have it set to "Allow Override of Predefined settings" for your nHancer's FSX PROFILE (as in Nick's setup guide)? If "yes", forget about any other same-type settings you "use" in FSX or the nVidia Control Panel. They ARE NOT being "used", and you should not be trying to use FSX itself or the nVidia Control Panel to "influence" them.5. If you are using nHancer to make an FSX Profile, but do NOT have "Allow Override of Predefined Settings"....huh? Why are you using nHancer in the first place? You use nHancer BECAUSE it provides MORE options than FSX or the nVidia Control panel to begin with for FSX. :(

[vgbaron 3,627]

Hm, I always thought NickN's guide is about optimizing system - neither performance nor quality, but inbetween. But I will check the setting of high performance though, to see if anything is to be gained from it. I use only Nhancer, solely for its 8xS setting, which is for me an absolute quality must.

You are correct. Nick has always maintained that he's looking for the best balance between all the parts of the system that give the best quality with the best performance. If one is ONLY interested in performance, I suggest putting all the sliders to the full left position and have at it. You'll have FPS in the hundreds, especially at FL30.And, that 'balance' is different for each system. that's why it is upsetting to some when someone posts a "do this and it will fix your FSX". Savvy users know better and might cautiously the "tweak of the week" but there are some newbies who take things as Gospel and have a really bad time.In this thread, the BP=0 is an awful mess FOR ME - and that's all that statement means. I tried it and didn't like it. For whatever reason, it upset the 'balance' of my system.IMHO, users are better off getting their system in a stable condition, the best thay can, and then slowly and methodically apply a tweak and test - tweak and test - there IS NO MAGIC BULLET.Vic

[jjjallen 37]

Can you give BP=3072 a try? I know that reading through 600 posts can be confusing, but I think it was mentioned that it gave the exact performance benefit of BP=0 without any glitches of any kind... I don't know if its true, because BP=0 is already stable for me, but It wouldn hurt if you try BP=3072and see if there is any difference.

I'm impressed. Better smoothness.Thanksjja

[Kosta 122]

At some rate, paralelly to the BP=0 and trying to achieve some results, with my PMDG/MD11 I have the known problems, and I went into some research, as I had these errors:Display driver stopped responding and has recoveredDisplay driver NVIDIA Windows Kernel Mode Driver, Version 197.13 stopped responding and has successfully recovered.I had this happen with all other drivers, so it's not a specific driver issue.Besides, I did read this here:http://forums.nvidia.com/index.php?showtop...amp;hl=nvlddmkmSpecifically:Graphics software vendors:

[SpiritFlyer 987]

Into the new machine.I suppose I should read the past few pages to do a catch up, but I don't want to take the time away from flying this incredible machine. There is a huge jump in performance over the Q9550. No problems, stutters, or disappointments yet. Smmoth and very fast! Put the old hard drive into the new machine, turned it on and away we go! :( Stephen

[Paul J 349]

Into the new machine.I suppose I should read the past few pages to do a catch up, but I don't want to take the time away from flying this incredible machine. There is a huge jump in performance over the Q9550. No problems, stutters, or disappointments yet. Smmoth and very fast! Put the old hard drive into the new machine, turned it on and away we go! :( Stephen

So ya got it! Welcome to the 4 gig club, Stephen! One heck of a difference isn't it! :( :( All the best,

[Noel 4,503]

So ya got it! Welcome to the 4 gig club, Stephen! One heck of a difference isn't it! :( :( All the best,

I tell you since lowering BP to 0, 1 or even 2MB, setting process priority automatically to HIGH, FPS_Limiter set at 30, I have truely flawless performance now, even on the 'old' Quad I have running. Not a hitch! I have good memory running at low latency settings for a non-onboard memory controller. I run nHancer at 8xs, 16x aniso with a display res of 1680 x 1050. I flew the wonderful PMDG J4100 in FTX Orbx Aus terrain with my usually nearly full right sliders, clouds out to 110nm, flawless smooth performance with no artifacts. Same same for PNW scenery. I have to thank NickN for making the case for better memory & timings for that memory. I changed to some good memory and a mainboard to support it in response to some comments he provided on this topic. I find in the most dense of scnenery I have to use SOME bufferpools over zero, 1-2mb seems perfect and gives me crystal clear scenery with no artifacts. Since making changes to memory and mainboard some time ago I have had very good performance. It seems the final tweaks to bring about flawless performance came from HIGH processs priority (automatically via Prio) combined with a specific low BP setting. I think at this juncture the main benefits from the latest CPU platform would be just more frames in the densist of sceneries and weather such as landing the PMDG 747 at JFK in weather. I'm sure that would threaten the rig I am on. I have not flown this bird for a while since making the changes described above, so will have to try that this weekend. Been enamored with the J4100, and the ERJ-145 and haven't been in the heavies since buying them and adding them to my FSPax airline. Pretty cool we keep finding useful tweaks in here. The do make a difference at times.I was looking back to see what Stephen was running in the way of memory with his 9550 because as I say exquisite performance is very possible with Core 2 quads as I can attest. I guess the sig gets updated retroactively.Noel

[SpiritFlyer 987]

Hi Noel,Indeed, this rig is better than I expected. I had a Q9550 @3.65ghz and 4g of 1066 ram on a Asus PK5-V MB with a GTX275-1792 ram card. On the ORBX forum I established some before and after screenshot benchmarks of several airports. Since I just ported my existing hard drive over to the new machine, the FSX setup was exact. The result was fairly comparing apples with apples between the Yorkfield and Bloomfield quads. I was really surprised to find an average of a 74.5% difference in favour of the newer machinery. Frankly I did not expect anything like that for my Q9550 based equipment was fast anyway. The numbers were so high I am suspicious and waiting for the other shoe to drop!You can compare the benchmarks RIGHT HERE IN THE ORBX FORUMS.. It's in the general discussion area titled "I am outta here (as into a new computer).By the way Bojote, I just tried the Bufferpools=0 with this machine and it was being strangled with artifacts. Has any progress been made toward a general consensus of the war parties as to the best settings yet or any other breakthroughs?Stephen

[Guest]

Has any progress been made toward a general consensus of the war parties as to the best settings yet or any other breakthroughs?

Set Texture filtering to High Performance instead of High quality: it makes a difference on my machine (not in quality, oddly enough, but only when it comes to artifacts!). But er... I don't think there is a general consensus yet on this one... :( (But er... afaik 'general consensus' and 'FSX' don't really go well together... :( )

[Noel 4,503]

The result was fairly comparing apples with apples between the Yorkfield and Bloomfield quads. I was really surprised to find an average of a 74.5% difference in favour of the newer machinery. Frankly I did not expect anything like that for my Q9550 based equipment was fast anyway. The numbers were so high I am suspicious and waiting for the other shoe to drop!Stephen

That's great, but as I say, already seeing liquid smooth at 30 frames pretty much everywhere inc FTX scenery--until very dense clouds come in, then it can drop off to the low 20's for me. I think your 1066 memory probably added enough incentive to move beyond what you had going previously.

[bojote 20]

By the way Bojote, I just tried the Bufferpools=0 with this machine and it was being strangled with artifacts. Has any progress been made toward a general consensus of the war parties as to the best settings yet or any other breakthroughs?Stephen

Stephen, for the past two weeks, I've been testing different scenarios, reading throught lots of posts and doing my own research. Also someone posted something that makes sense and confirms what I've seen so far, but until people test it there is nothing conclusive about it.Do this VERY SIMPLE test, DISABLE aniso in nHancer (yes, you read right) the option that says OFF (not application) OFF.. ( and don't worry, it WILL NOT actually disable it as long as FSX has the anisotropic filter option enabled inside the sim) thenm tell me if you still see artifacts.. just do that simple change. I'll elaborate further, but I need you to test that.

[David Roch 1,982]

Almost 10 hours flying now with no more artifacts.Smooth and crispy as you can see:Here are my settings if it can help:1) NO nhancer!2) Settings in Nvidia CP as following (red rings show important settings!)3) FPS limiter @254) FSX.cfg only tweaks:-TEXTURE_BANDWIDTH_MULT=100- poolsize=0(no AffinityMask!)5) Display settings:Average 25 fps in PNW non-urban areas with driver 196.21.

[Guest Shockwave]

Stephen, for the past two weeks, I've been testing different scenarios, reading throught lots of posts and doing my own research. Also someone posted something that makes sense and confirms what I've seen so far, but until people test it there is nothing conclusive about it.Do this VERY SIMPLE test, DISABLE aniso in nHancer (yes, you read right) the option that says OFF (not application) OFF.. ( and don't worry, it WILL NOT actually disable it as long as FSX has the anisotropic filter option enabled inside the sim) thenm tell me if you still see artifacts.. just do that simple change. I'll elaborate further, but I need you to test that.

What WILL change is that in the driver / enhancer you can select what level of Aniso you want from 4x thru high quality 16xAF. If you let FSX chose it wont be 16XAF (8X and lower depending on generation of card). If you use high resolution screen sizes, lower quality Aniso filtering shows up easily, if lower res - not as much.Also linked to AF is when you select "texture quality" settings in the driver/nhancer. Settings of Higher or mid performance uses AA/AF and texture LOD performance optimizations (shortcuts) however when you chose high quality settings those optimizations are all turned off.Want less spikes less black screens - just ignor nhancer/driver values of higher quality AA and use FSX AA (2xAA) which will give you near DX10 performance and not use as many resources as high quality settings. Basicalay more filtering=less perfomrance vs what your hardware can do vs. what you can live with.Edit: Nice Pic David! Notice your Agen is at "normal", if you bump it up to dense what happens? Just be aware without using nhancer your AA in FSX is not what you have selected in the NV CPL and defaults to stock 2xAA a bit like in DX10 preview mode, been broken that way for sometime.

[bojote 20]

Just be aware without using nhancer your AA in FSX is not what you have selected in the NV CPL and defaults to stock 2xAA a bit like in DX10 preview mode, been broken that way for sometime.

Hi David,Remember AA can also be tweaked inside FSX... [GRAPHICS]MultiSamplesPerPixel=8MultiSampleQuality=2