I think you'll find LM backtracked from that, and for good reason. Also it mentions nothing of HT. It takes less than minutes to establish the correctness of what I have stated about the AM with simple testing.

 

 

 

Carlo,

Just do a fps test with AM added +4 fps i get by adding it.



Worth a try, nothing to lose apart from 2 mins of your time

 

Fair enough. I will give it try.

Fair enough. I will give it try.

It is a good idea to monitor the LPs loading up the sim threads, with the Task Manager "Performance" tab.

I can also confirm the big benefit of this tweak if I use a PMDG-plane, thank you very much for this!

 

But very interesting: It doesn't work with the iFly 747V2 - still bad frames.....

Excellent find!!!

I’ve just completed some rudimentary tests on my system. For the first batch of tests I used UK2000 EGLL extreme with clear skies and my Long Haul Config which is similar to the one Rob posted. I also set the frames to Unlimited and measured the average FPS using FSUIPC over a set time for each run.

Default Bonanza = 3% Increase in FPS
PMDG 737 = 5% Increase in FPS

 

I then switched to LOWI which on my system is a default airport but used my heaviest settings together with ASN
PMDG 777 = 8% increase in FPS

 

Conclusion: Based on my system, the more complex the aircraft the better the results.

All these tests were stationary so it doesn’t show an overall picture of what you might experience on a set flight.

 

Regarding VAS, there is a tiny difference and you do use some which is to be expected, but the numbers are small. I measured 19K more VAS usage on my 777 test. This number is insignificant compared to the benefit. I used FSUIPS to measure.

 

OPTIMIZE_PARTS=0 Monitor IPC:024C (S32) = 1118504
OPTIMIZE_PARTS=1 Monitor IPC:024C (S32) = 1084688

Are you guys testing this with you AI turned on? Can you state to what level is your AI set to ?

Did some tests in my standard testing scenario (PMDG 777, Aerosoft EDDF at holding point of runway 07R, same date and time and weather in all cases):

 

AM=255 (4770K, all cores on), OPTIMIZE_PARTS=0 ----> 34.4 fps, 8% fps variance

AM=84, OPTIMIZE_PARTS=0                                     ----> 36.2 fps, 8% fps variance

AM=255, OPTIMIZE_PARTS=1                                   ----> 46.5 fps, 8.5% fps variance

AM=84, OPTIMIZE_PARTS=1                                     ----> 49.0 fps, 8% fps variance

 

FPS numbers are average fps after 3 minutes in the VC. Both tweaks (AM, OPTIMIZE_PARTS) have no impact on VAS. Even though the AM=84 improves fps a bit, I'm not sure if it introduces slight stutters and slower texture loading. Have to do more testing on that. But the OPTIMIZE_PARTS tweak is definitely a keeper! Great find!

It seems people are seeing the biggest improvements with a PMDG aircraft am I right? I don't have anything PMDG for P3D, just their NGX in FSX. However I took the hardest hitting aircraft I have for P3D, the Aerosoft Airbus bundle and dropped in at ORBX KPSP. With the setting left at 0 I had 21-23 FPS and with the setting at 1 I had 23-26 FPS. So a slight increase there as well.

LM has not backtracked on that, unless you can find an official quote.

Did some tests in my standard testing scenario (PMDG 777, Aerosoft EDDF at holding point of runway 07R, same date and time and weather in all cases):

 

AM=255 (4770K, all cores on), OPTIMIZE_PARTS=0 ----> 34.4 fps, 8% fps variance

AM=84, OPTIMIZE_PARTS=0                                     ----> 36.2 fps, 8% fps variance

AM=255, OPTIMIZE_PARTS=1                                   ----> 46.5 fps, 8.5% fps variance

AM=84, OPTIMIZE_PARTS=1                                     ----> 49.0 fps, 8% fps variance

 

FPS numbers are average fps after 3 minutes in the VC. Both tweaks (AM, OPTIMIZE_PARTS) have no impact on VAS. Even though the AM=84 improves fps a bit, I'm not sure if it introduces slight stutters and slower texture loading. Have to do more testing on that. But the OPTIMIZE_PARTS tweak is definitely a keeper! Great find!

For a 6 core with what would you recommended

Going to try it as well.

Anyone tried it with the RealAir Turbine Duke and can confirm it works? I'm most interested in heavy cloud situations where FPS drop in the Duke.

Even though the AM=84 improves fps a bit, I'm not sure if it introduces slight stutters and slower texture loading.

 

 

Which is why one shouldn't bother with an AM. The logic behind 84 is to make sure that P3d's main thread is prevented from using physical core 0. There is some urban legend that "the OS runs on physical core 0" which hasn't been even close to being true since the days of Win XP. The same is true of using AM=14, with HT off (HT should be off, unless one can't change the setting in the BIOS).

 

If your system is being used properly, there should be very little else that is requiring CPU resources while P3d is running. That doesn't mean that all CPU cores will be running at 100%. It just means that if one looks at the task manager, P3d will be using 99% of the CPU resources that are being used in toto by all processes.

Anyone tried it with the RealAir Turbine Duke and can confirm it works? I'm most interested in heavy cloud situations where FPS drop in the Duke.

 Yes, OP=1 does work with the Turbine Duke for me.

Sim stopped: average frame rate for last   1425 secs = 67.1 fps   up from 61.3 fps.   Carenado TBM 850. Going to test in heavier Weather but an 

increase for sure. 

Can any of you old school dual install chaps confirm if this also works with FSX? 

I did some tests with my system (i7 3930K @3.2GHz, GTX Titan, 16GB RAM) using the following aircraft at EDDH (German Airports):

 

                            Optimize_parts=0                  Optimize_parts=1

PMDG 737             16 - 17 fps                              18 - 19 fps

 

iFly 737                  14 - 15 fps                              15 - 16 fps

 

AS A320                 17 - 18 fps                              18 - 19 fps

 

 

One could say that the tweak gained 1 or even 2 fps on these aircraft - but that is all. I therefore can not share the great enthusiasm!

 

I also did a flight with the PMDG 737 to LSZH but when reaching FL330 (I admit of playing a bit with the simulation rate) my P3D 12944.0 just froze, something that did not happen since 2.2 (or was it 2.3?). Have to do some more testing to find out what happened here and whether it is connected to the tweak or not.