Preload radius don't affect RAM usage, require restart

[BJ55]

As per Manual: "A larger setting reduces slowdowns due to hard disk swapping, but requires more RAM." , but if I set preload radius to 10000 or 150000 the memory usage is the same:
F-16C Afghanistan C&D instant action





Also, the preload radius setting is not applied without restarting DCS, logs sequence:
Started DCS, changed preload from 80k to 150k= "dcs-01 150 previous 80.log"
Restarted DCS                                                     = "dcs-02 150 restarted 150.log"
Started DCS, changed preload from 150k to 10k= "dcs-03 10 previous 150.log"
Restarted DCS                                                     = "dcs-04 10 restarted 10.log"

Since recently some users had issues running DCS as admin, notice that I use an Admin account and everything was installed as Admin from day one, so I cannot perform tests with a UAC enabled Standard account.

dcs-02 150 restarted 150.log dcs-03 10 previous 150.log dcs-04 10 restarted 10.log dcs-01 150 previous 80.log

Edited April 6 by BJ55

[MAXsenna]

11 minutes ago, BJ55 said:

Since recently some users had issues running DCS as admin, notice that I use an Admin account and everything was installed as Admin from day one, so I cannot perform tests with a UAC enabled Standard account.

If you are logged into the Windows admin account it's not applicable, as that account has access the administrator profile. The issue is when a user is logged in with another account which doesn't have access to the administrator profile. 

I don't use preload radius at all, it's better on my system.  

[Hiob]

Main difference I notice is that the initial load time to get in the cockpit is much longer with a higher preload radius.

Unless you notice any streaming/loading hickups while flying, I would try with a small preload radius.

[Mr_sukebe]

Go have a check on VRAM usage changes (use, not reserved).

Thats now visible on the FPS counter option in game

[BJ55]

Both RAM/VRAM usage are the same, tested with ReBar on and off.

[Czar66]

4 hours ago, MAXsenna said:

I don't use preload radius at all, it's better on my system.  

Hi, Max. How is that? You keep the preload radius on minimum?

[Qcumber]

I did a test comparing performance with preload at 60 vs 150. There was a small difference in GPU latency and about a 1Gb difference in VRAM use. 

 

 

[Mr_sukebe]

2 hours ago, BJ55 said:

Both RAM/VRAM usage are the same, tested with ReBar on and off.

 

That's interesting.  Haven't checked VRAM changes in my own system for that in months.  I'll go have a scan later on.

[MAXsenna]

Hi, Max. How is that? You keep the preload radius on minimum?

Yup, I was gonna test after I fixed my issue with not being able to run DLAA, and I started with min minimum settings and never touched the prereload settings, and was running better than before and way less load times.

Sent from my SM-A536B using Tapatalk

[BJ55]

I would be glad to know if other users noticed the same behaviour of the preload setting not being applied, if dev's are aware of a issue/bug or if it's happening only on my PC.

I see lots of FPS drops/frame time spikes/sutters reports, this is what Process Monitor logs during the spike:

DCS is constantly reading from *.surface5, even when parked, as if preload radius is zero.