Bufferpool settings

[BogeyFree 0]

I just switched over to DX10 using Steve's Fixer and followed all the recommened settings for NI and fsx.cfg but I have no idea what to set for the Bufferpools.  Looking at my PC specs can somebody suggest a good starting point for both Poolsize and RT?  Thanks.

[Paul J 349]

Read the How-To and the BP doc, Bogey. Only you can make a starting point and then move upward or downward from there. Set up a "heavy" test circle/circuit (traffic, clouds, etc) using the default C172, in or around an area like LosAngeles LAX, London, Vancouver, Seattle, etc, start and fly it at a constant altitude - say 1000, 1500, maybe 2000', in the air and then save it. Set up your BP based upon your proc speed and your GPU speed. Then fly exactly the same course five or six times, noting the trend of your fps. This is how to tweak the BP setting.

 

       The questions are 1). Is my (4.3gig) CPU better/faster than my (GTX460) GPU? If yes - then UsePools=1. If no - then UsePools=0. If it's a 1, then you need to add a RejectThreshold setting and test again, and again, using up to six different settings (see the guide). If the performance goes down with an RT, then try UsePools=0, and then test again.

 

       This one can go either way, but I would try UsePools=0 first, as that is easiest, and will yield the highest performance. Once you've got a decent idea of fps in the circuit - try it again with UsePools=1, RejectThreshold=524288. This is big enough that it will either lower the performance, or it will increase the performance. The BP doc gives you more figures, and a fair idea of how bp's work.

 

The system looks ok, for sure, the single monitor is good, as the proc is low-end and tied to a low-end gtx460, so BP might go either way, but could yield decent performance because of that monitor. Much will depend upon stopping processes and services, and how you setup your Inspector - which is going to determine your AA and shimmering. This is then going to be modified by your own expectations of what is visually acceptable!

 

Well, I hope this helps!

 

All the Best,

 

pj

[BogeyFree 0]

Ok thanks Paul.  I'm out of time now and will have to resume taking a look at it when I get back from my vacation.

[PapaRomeo 40]

Sorry, but I'm not trying to hi-jack this thread, just one question.

 

I just recently got a GTX780 with 3GB ram.
Should i use Bufferpools=0 ?

[Michael Moe 497]

Sorry, but I'm not trying to hi-jack this thread, just one question.

 

I just recently got a GTX780 with 3GB ram.

Should i use Bufferpools=0 ?

Unless you have a 5ghz Haswell i would say yes you should.

 

Michael

[Paul J 349]

A 2600K at 4.4?? - are you kidding ???  of course you should.  "UsePools=0"

 

The 780 is w-a-a-a-a-a-a-y faster than a slow-poke 4.4-gig proc. Please - listen to Michael - that 780 rocks, big-time!!

 

...and why are you "babying" your (beautiful) 2600K?   :lol:  There are people on this forum that would cut off their b right arms to get their hands on a 2600K/2700K processor!

 

i7-26-2700K's kick butt man!! They don't need de-lidding... and they're easy to cool, and they are so easy to overclock.... 

 

pj   :biggrin:

[Jedi928 6]

Hi Paul,

 

And with i7 2600 K 4.4 and a 580 GTX 3G what do you think BP=0 or 1   

580GTX is faster than CPU ay 4.4 or not?

[messifan10 3]

i have testet my i7 3770k 4,5 and a gtx 780 i see NO diffence at all with bufferpools on or off !!!

if you ask me it dosent matter !

[Paul J 349]

Hi Paul,

 

And with i7 2600 K 4.4 and a 580 GTX 3G what do you think BP=0 or 1   

580GTX is faster than CPU ay 4.4 or not?

Yes - 4.4 is slow. The 580 will eat it for breakfast. It's also "UsePools=0", not "BufferPools=0". I'm not 1000% sure that that is the case, but "BufferPools=0" (or 1) doesn't make for logical naming convention sense. "UsePools" is certainly correct, tested - and works.

 

 

i have testet my i7 3770k 4,5 and a gtx 780 i see NO diffence at all with bufferpools on or off !!!

if you ask me it dosent matter !

Then you didn't do your testing correctly, or your syntax was incorrect.

 

pj

[jayfresh 0]

Yes - 4.4 is slow. The 580 will eat it for breakfast. It's also "UsePools=0", not "BufferPools=0". I'm not 1000% sure that that is the case, but "BufferPools=0" (or 1) doesn't make for logical naming convention sense. "UsePools" is certainly correct, tested - and works.

 

 

Then you didn't do your testing correctly, or your syntax was incorrect.

 

pj

Hi paul i am using i5 4670k overclock 4.3 n  evga 660 non ti sc signature 2 what setting should i use BP=0 OR BP=1 with RT?

[Jedi928 6]

Thanks Paul,

 

for sure it is

 

[bufferPools]

UsePools=0

[messifan10 3]

syntax ? or testing correct ?

[Paul J 349]

Hi paul i am using i5 4670k overclock 4.3 n  evga 660 non ti sc signature 2 what setting should i use BP=0 OR BP=1 with RT?

 Hmmm... Could go either way, Jay, so create and save your test circuit first - then set the buffer to "UsePools=0" first - because that's the easiest: do five or six circuits noting the fps, then change it to UsePools=1, with an RT of, say 524288 (512 kB). Repeat until you have the highest and most consistant rate. Your TBM should probably be set to 80.

 

Take your time.

 

All the Best,

 

pj

[SteveW 1,792]

 

 

Repeat until you have the highest and most consistant rate

Hi Paul,

Is there a program that can measure the average fps of FSX/P3D over a set period or something like that?

regards

steve

[jayfresh 0]

 Hmmm... Could go either way, Jay, so create and save your test circuit first - then set the buffer to "UsePools=0" first - because that's the easiest: do five or six circuits noting the fps, then change it to UsePools=1, with an RT of, say 524288 (512 kB). Repeat until you have the highest and most consistant rate. Your TBM should probably be set to 80.

 

Take your time.

 

All the Best,

 

pj

Thank u paul will try both to see which gives me the highest consistant rate also if u do not mind can u give me an example with the BP=1 WITH RT PLS