hotstick

Hello Flappie, were you able to try and reproduce the problem?

Even though DCS performs its analysis at launch, it's still possible to disrupt its operation by modifying the core affinity on the DCS task using the task manager (DCS is already running). This isn't a permanent action, but it at least allows you to run tests (excluding process lasso) and see if stuttering is still present.

The YouTuber at the video link below describes the goal, which is to have the DCS game on the core with cache and everything else on the core without cache. He also mentions the fact that, as it stands, Windows management (and perhaps also DCS) doesn't allow this goal to be achieved. However, it is possible to achieve this by manually addressing the problem (process lasso and other actions). It's worth a try!

Thank you for all this information and for the dcs.log. Here's the dcs.log file for a 14900K, which confirms the same warning message at the beginning of the log. === Log opened UTC 2025-03-21 03:56:55 2025-03-21 03:56:55.110 WARNING EDCORE (Main): CPU HAS PARKED LOGICAL CORES (this can be source of stuttering and reduced performance especially on hybrid CPUs with P/E-cores) 2025-03-21 03:56:56.563 INFO EDCORE (Main): Platform Timer Resolution: 15.45ms 2025-03-21 03:56:57.191 INFO EDCORE (Main): Created boot pool: n:32 2025-03-21 03:56:57.192 INFO APP (Main): Command line: "C:\Program Files\Eagle Dynamics\DCS World\bin/DCS.exe" 2025-03-21 03:56:57.192 INFO APP (Main): DCS/2.9.14.8222 (x86_64; MT; Windows NT 10.0.26100) 2025-03-21 03:56:57.192 INFO APP (Main): Application revision: 248222 2025-03-21 03:56:57.192 INFO APP (Main): Renderer revision: 25901 2025-03-21 03:56:57.192 INFO APP (Main): Terrain revision: 25769 2025-03-21 03:56:57.192 INFO APP (Main): Build number: 305 2025-03-21 03:56:57.192 INFO APP (Main): CPU cores: 24, threads: 32, System RAM: 65298 MB, Pagefile: 9216 MB 2025-03-21 03:56:57.192 INFO APP (Main): CPU info: system affinity mask: 11111111111111111111111111111111 process affinity mask: 11111111111111111111111111111111 CPU: Intel(R) Core(TM) i9-14900K [1x L3 caches] CPU has uniform cache size 36 MB (per cache) CPU cores have different efficiency classes: [0-1] logical cores with efficiency class 1: {0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15} logical cores with efficiency class 0: {16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31} CPU cores have different performance classes: [0-2] logical cores with performance class 2: {8, 9, 10, 11} logical cores with performance class 1: {0, 1, 2, 3, 4, 5, 6, 7, 12, 13, 14, 15} logical cores with performance class 0: {16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31} common cores: {4, 5, 6, 7, 12, 13, 14, 15} render cores: {8, 9, 10, 11, 0, 1, 2, 3} IO cores: {16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31} unavailable cores: {} However, a 12700K which is "a little" hybrid with its 4 e-cores does not have the same alert message at the start of the log: === Log opened UTC 2025-04-11 12:15:32 2025-04-11 12:15:32.826 INFO EDCORE (Main): Platform Timer Resolution: 5.85ms 2025-04-11 12:15:32.899 INFO EDCORE (Main): Created boot pool: n:20 2025-04-11 12:15:32.900 INFO APP (Main): Command line: "D:\j\Eagle Dynamics\DCS World OpenBeta\bin-mt\DCS.exe" --force_disable_VR --restarted 2025-04-11 12:15:32.900 INFO APP (Main): DCS/2.9.14.8394 (x86_64; MT; Windows NT 10.0.22631) 2025-04-11 12:15:32.900 INFO APP (Main): Application revision: 248394 2025-04-11 12:15:32.900 INFO APP (Main): Renderer revision: 25901 2025-04-11 12:15:32.901 INFO APP (Main): Terrain revision: 25970 2025-04-11 12:15:32.901 INFO APP (Main): Build number: 306 2025-04-11 12:15:32.901 INFO APP (Main): CPU cores: 12, threads: 20, System RAM: 65367 MB, Pagefile: 4096 MB 2025-04-11 12:15:32.901 INFO APP (Main): CPU info: system affinity mask: 11111111111111111111 process affinity mask: 11111111111111111111 CPU: 12th Gen Intel(R) Core(TM) i7-12700KF [1x L3 caches] CPU has uniform cache size 25 MB (per cache) CPU cores have different efficiency classes: [0-1] logical cores with efficiency class 1: {0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15} logical cores with efficiency class 0: {16, 17, 18, 19} CPU cores have different performance classes: [0-1] logical cores with performance class 1: {0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15} logical cores with performance class 0: {16, 17, 18, 19} common cores: {8, 9, 10, 11, 12, 13, 14, 15} render cores: {0, 1, 2, 3, 4, 5, 6, 7} IO cores: {16, 17, 18, 19} unavailable cores: {} I won't explain the reason, just to provide food for thought. Regarding the 9950x3d, the hardware allows for core separation between the game and the rest, but optimizations and tweaks are required to achieve this result. Can ED confirm that they're working on this issue? A little more transparency on how core processing currently works and future developments would help us move forward. In the meantime, perhaps Process Lasso could clean things up. Has anyone tried it?

If ED would be kind enough to give us some explanations, that would be great.

It's a significant advantage that DCS doesn't park non-cached cores. This allows the DCS game to occupy the 8 cached cores and leave all other processes with the non-cached cores. This way, DCS has all 8 cached cores to itself, unlike a 9800x3d, which has to share some of its cores with other processes. The 9950x3d shares the same advantage of the 14900 (many e-core) over other processors in the same family. The 1% and 0.1% are improved. Part of a dcs.log on a 9800X3D. common cores: {8, 9, 2, 3, 12, 13, 14, 15} render cores: {0, 1, 4, 5, 10, 11, 6, 7} IO cores: {} unavailable cores: {} Can someone with a 9950X3D post the same part of their dcs.log?

Could you make your DCS log file available? I'll compare it to the 9800X3D log from a core allocation perspective. Thanks.

This video is extremely important! I have a question about FPS. Despite the much lower latency of the 9950x3D, the FPS in games are identical! Do you have any idea why?

Hello, here is the script in Export.lua And the result in the file "C:\Users\xxxxxxx\Saved Games\DCS.openbeta\Logs\TargetHotStick.log" The problem is that when I enter a Tank or something else using Combined Arms, the TargetHotStick.log file is always empty.

I created a utility that allows you to launch or change the profile (Target TMC) of a Thrustmaster Warthog hotas while the DCS game is running. To do this, I capture the name of the current module via DCS export.lua using "LoGetSelfData()" and property "Name". The two functions used to receive the module name are: -LuaExportStart() -LuaExportOnHumanStart() This works fine except for "Combined Arms" or "tactical commander" because this module isn't detected by the two functions used. Is this normal? Thanks.

Thanks Morpheus for the answer I didn't expect to search for it under this term. Plus the directory already existed (I'm not smart).

I can't find the directory name for kneeboard in \Saved Games\DCS.openbeta\Kneeboard I tried: OV-10A OV-10 OV10A OV10 Without success. Do you have the correct directory name? Thanks.

Moire with DLSS (not 4) nevertheless remains a real problem

Oui les relations pour le moins litigieuses ou conflictuelles qui relient ED et ses sous-traitants ne devraient pas impacter les clients (d'ED). Les évènements récents (ou plus anciens) dans ce domaines impactent notre plaisir à voler sur votre simulateur. Je pense que vous devez le savoir mais aux yeux des clients ce type de problème n'aura quasiment aucune retombée sur les sous-traitants, mais bien sur ED. Si je peux comprendre qu'il puissent y avoir certains différents entre sous-traitants et ED (la vie d'une entreprise n'est pas un long fleuve tranquille), c'est bien à l'entrepreprise en titre "ED" de protéger ses clients. J'espère que vos méthodes ou relations avec vos sous-traitants seront adaptés pour protéger vos clients (en respectant aussi les sous-traitants lorsqu'ils le méritent). Longue vie à la simu (et on compte sur vous "ED" pour cela)

On the other hand, it is very blocking for those who fly in an VR. Indeed being in flight (in VR), Combined Arms becomes unusable. You would have to exit DCS and relaunch in 2D to have Combined Arms fully operational.

It would be truly immersive if VR was implemented. I can not wait to see him.

I had both versions of DCS installed. I haven't updated one of them and pointing OpenKneeboard to the one that hasn't been updated, OpenKneeboard works normally. It would certainly be possible to isolate the files necessary for OpenKneeboard so that it works on a "Fake" of DCS.

I have the F-15E, the Mirage F1 and the Apache and it crashes. I don't have the MB-339. I also removed all the mods I had and it continues to crash.

I also use it commonly and if the author can no longer maintain this application it's a disaster.

Same error ! --------------------------- There was an error! --------------------------- Use Ctrl + C to copy this error and e-mail it to s13solutions@gmail.com : System.ArgumentNullException: La valeur ne peut pas être null. Nom du paramètre : path à System.IO.Directory.GetParent(String path) à KneeboardBuilder.Viewmodels.MainViewModel.BackupAircraftDefaultGroups(String _origPath) à KneeboardBuilder.Viewmodels.MainViewModel.LoadAircraftData() à KneeboardBuilder.Viewmodels.MainViewModel.Initilization() à KneeboardBuilder.Viewmodels.MainViewModel..ctor() à KneeboardBuilder.Views.MainView..ctor() --------------------------- OK ---------------------------

And not on Fw-109K4

I have the problem on F-16C but not on A-10C II (with TGP activated)

Flash for me too. Context: Latest Nvidia driver no DLSS MSAA x 2 Pimax Crystal Flight on F-16C

I have a problem with Crystal Palace, impossible to hear. On several different solo missions. However, I do solicit it on COM1 (F/A-18C) on the frequency 142.000 Voice Attack tells me: TX1 COMM1: ARC-210 AM 142.000 Mhz tuned for auxiliary unit Crystal Palace [AI] then [Read mission Briefing] then Constructing message ... then Done But no audio feedback. Do you see a reason ? Thank you.

Oui c'est effectivement une bonne idée car du coup la "souris" est sensée se trouver toujours au même endroit. Et il en faut deux (gauche et droite) car des fois une main peut se libérer et pas l'autre ! @Valerian73 Je n'ai pas encore reçu le produit. Il les fait lui même et c'est long. Pour revenir à mon explication sur la visée de la souris à la tête ça fonctionne bien et c'est plus efficace pour un ou deux clics que de prendre la souris.