One can also use such a converter to get the right values for AM.

http://www.mathsisfun.com/binary-decimal-hexadecimal-converter.html

Spirit

 

 

This is certainly true while CPUs have low core counts.

 

Do mean number of cores or number of threads?

I think he meant to say that 'this is true if a processor has a low core count' ie 2 cores or 4 threads.

Then you will see, in general, better performance by using all cores/threads.

I added AM=15 from having no entry.  On my i5 [email protected] it raised the temp by 6 degrees to 66.  My GPU fan seems to be working harder on my GTX580 but I can live with this.  The micro-stutters have reduced significantly.

One (compound) word for setting affinity "ProcessLasso". Works great. Here is how I am using with FSX: http://uchisworld.wordpress.com/2014/06/15/malleus-maleficarum/

 

I found this software very useful and now have it running with P3D. Seems to do the job of keeping the experience smooth. 

 

My head started hurting trying to work out  this affinity mask stuff so I just use ProcessLasso. Seems to work well. Any others tried it?

 

Adam

Thanks Toby, will give it a try.

I found this software very useful and now have it running with P3D. Seems to do the job of keeping the experience smooth. 

 

My head started hurting trying to work out  this affinity mask stuff so I just use ProcessLasso. Seems to work well. Any others tried it?

 

Adam

 Adam,

 

You may want to do as Toby23 suggested and not use the thread priority part of ProcessLasso, depending on your rig. But the affinity settings part is MUCHO easy.

Do mean number of cores or number of threads?

Cores.

 

It seems reports indicate that on an Intel 4 core CPU AM=15 (1111) is better than default. So perhaps it is true that the LM default is AM=14 (1110), and that P3D is better with 4 cores than 3. However, as we add more cores, there will be a point where splitting the app over all those cores becomes unimportant, or even inefficient.

Cores.

 

It seems reports indicate that on an Intel 4 core CPU AM=15 (1111) is better than default. So perhaps it is true that the LM default is AM=14 (1110), and that P3D is better with 4 cores than 3. However, as we add more cores, there will be a point where splitting the app over all those cores becomes unimportant, or even inefficient.

 

I continue to marvel at the belief that the LM authors got this wrong...

 

Maybe each user should do what he or she feels works well on their system without generalizing

this to be a universal remedy..  <_<

Cores.

 

It seems reports indicate that on an Intel 4 core CPU AM=15 (1111) is better than default. So perhaps it is true that the LM default is AM=14 (1110), and that P3D is better with 4 cores than 3. However, as we add more cores, there will be a point where splitting the app over all those cores becomes unimportant, or even inefficient.

 

Why in that case, does P3D  run better only using 75% of its available processers?

And that's exactly why I posted this question over at LM, in the hope that they can provide an answer.

The majority of you seem convinced that running on all cores works best. It doesn't.

If you only have a dual core or an older chip with no hyperthreading, then using all cores seems to be the best solution.

http://www.prepar3d.com/forum-5/?mingleforumaction=viewtopic&t=6380.4

And that's exactly why I posted this question over at LM, in the hope that they can provide an answer.

The majority of you seem convinced that running on all cores works best. It doesn't.

If you only have a dual core or an older chip with no hyperthreading, then using all cores seems to be the best solution.

http://www.prepar3d.com/forum-5/?mingleforumaction=viewtopic&t=6380.4

 

According to your link Beau Hollis (Rendering System Lead - Prepar3D® Team) on on: April 16, 2014 said:

 

Our default AM masked off the first core because some OS tasks happen on the main core and in the past we have seen more stable results masking it off provided you have a 4 or more cores to play with. As I said a little earlier in this thread, we are changing that for 2.3 to use all cores and will continue to look into paging performance as a whole.

 

 

that suggests:

 

a - AM 15 (1111) is better for less than 4 cores, and

 

b - v2.3 will use all cores

I hope that this will all become irrelevant when LM release 2.3 and Intel release their new 6 core chips in a few months.

This is all well and good but have you guys actually even tried just using default? The devs have said it multiple times.  They changed the way processor cores are detected and used quite a while ago (at least since 2.0). Most intel chips will be detected and used properly without this tweak at all. Usually the only people that need this are those using AMD chips that do not get detected properly.

 

On my I7 6 core/12 thread it runs exactly the same as if I were to set a AM=1364.

 

Another post from one of the devs I read earlier this week also said that in 2.3 the default behavior will be to enable ALL cores because it runs smoother.

Have you tried reading the first post from beginning to end? or any of the posts in this thread?

The whole reason that the flight sim community has it's knickers in a twist is because, for many people, P3D does not run optimally using default settings.

 

With an i7 quad core with default settings, I get blurry ground textures and stutters in turns.

BUT with a custom affinity mask value, I have the smoothest and sharpest flight sim experience I have seen in nearly six years.

 

If you had read any of these posts, you might also have learnt that 6 core appears to provide the optimum environment for P3D, meaning you won't have to tweak and can fly with default settings. Now get back in your cockpit and enjoy your 6 core flight sim experience.

 

... have you guys actually even tried just using default? ..