BJ55

From the resmon graphs it seems to me that any program tested is using perf2 cores for rendering, perf1 cores for rendering/common and perf0 for IO, the only one with a stuck core while idle is DCS. (be aware that my W10pro installations are highly customized/debloated for performance and with updates disabled after last stable) I'm so lucky that the HW required for my CAD-CAM is not compatible with W11, so for at least 2y I will keep W10 on all WS/PC, EOS doesn't mean that it will stop working the day after, it takes a total rewrite of DCS core engine to cause potential issues. (hate W11 UI, instability and new spyware) As I said MSI AB is used only for fan control (with 50-54°C, without 65-70°C), I always use only NVCP for change settings, if needed. Avoid ASUS sw like the plague, in particular Armoury Crate or any other phoning home "feature" installed in the BIOS. Totally agree, my next machines will be all AMD.

I'm curious about this, overall I have a good performance in SP, except in high density areas or with a huge number of units. I only use MSI AB for fan control, custom balanced energy plan and GPU at prefer max. perf. Do you think I need PL? The cores stuck at 100% is a regression that will be fixed with the next patch? dcs.log DxDiag.txt

Not a shader issue, there are too many objects to render, if you add all the computations for systems and units you reach the point where the hardware can't handle anymore, AH-64 @ Baghdad is even worse (512, med textures, no shadows, LOD at 0.8, 2x MSAA or DLSS, with VID feed = 40-50 FPS). This is a general issue and I don't expect both eye-candy, full-fidelty systems and performance at the same time... Maybe some more code optimization and add an option to reduce scenery objects? (quantity, not LOD)

Something's missing in your power options: You can also download the freeware Bitsum ParkControl and turn off parking for both P E cores. In Settings disable "start at login", once configured there's no need to run the program in background.

I know that (sadly) it's bugged and unmaintained, but at least the PCM meaning is explained in the manual.

The rudder is inverted.

I had problems at Kandahar too, after some tests I've found that performance issues are mission related, below the comparison with landing at Bagram from a TTI mission and from a ME custom mission with only 2+2 units: Later today I will test a complex ME mission. Edit: confirmed, on my PC multiple units can cause massive stuttering above Kabul.

Plenty of similar reports here: https://www.nvidia.com/en-us/geforce/forums/game-ready-drivers/13/557655/geforce-grd-57247-feedback-thread-released-22025/ , install previous drivers, don't install NV app. 552.44 is OK, not perfect like 314.22 but...

@Aero4000 : have you tried with no axis deadzones in controls settings? I had the same sensations until I've set deadzone to 0 and curvature to 15, firmware deadzones are enough, now it's smooth and predictable. Deadzones caused near center lack of precision with other aircrafts too.

Here's the mission/track/screenshots. Starting point-IP 10nm, IP-TGT 2.5nm, put in active pause to fill the bombing calculator and set the weapons system, hit shift-F10 once to hide labels, turn off active pause, stabilize the aircraft for the run. F-4_ag_train_cauc_82SE.zip

Thanks. At least now I know the cause of DCS DNS queries.

@silverdevil : I've wrote that DCS is not using the Windows DNS Client service, look at the second pic, no svchost.exe/DNS requests after flushdns. (btw: started with Basic in '85, does that count?)

I use Quad9 because my ISP is unreliable (I don't trust Google/Cloudflare). If a program allowed to TCP at some IP's suddenly tries to UDP then I recieve the request above.

No, quad9 is in my IPV4 settings and no, I never wrote that the DNS is hard coded, only that DCS is making request to the DNS server, instead of using the Win DNS Client, like all my installed software does. For the hardcoded DNS: what about Google devices OS/SW?

One of the two... Unfortunately I'm unable to share a video and the tracks have issues (see https://forum.dcs.world/topic/361617-jester-speed-call-out-on-loop/#findComment-5547877 ) , I have a screen took after a bomb run: When I don't use my corrections I follow Aero4000 suggestion: https://forum.dcs.world/topic/357226-laydown-cbu-87-falling-long/#findComment-5563802 (there you can find the simple training mission I'm using).

Why, after the update, DCS is not using Windows DNS Client and instead is connecting directly to the configured DNS server?

Thanks for the info Zabuzard, better if I read again the Aircrew Weapons Delivery Manual... The problem is that If I release when above the IP the 82SE high drag fall long, that's why I must subtract 10mils from the bombing computer value (and start releasing before reaching the IP).

Should or may? Is Laydown mode simply a Direct mode + a timer that starts when weapon release button is depressed? By the way, I'm still having the same bad results, even when flying with precision the profile, then I started to make my own "add-subtract" tables, for example L Mk-82SE 400TAS 1000ft 2.5nm 44300lbs: -10mil (64mil instead of 74).

Look at this:

Who's gonna take the workload of checking if files are safe before they're uploaded in the DCS updates delivery server? You know, a lot of users have to make AV exclusions and some missions/campaigns need desanitizing, immagine if a bad actor steals the credentials of a contributor...

Rise the nozzles.

Without your system info and event description it's really difficult to diagnose all the issues. CCXProcess (Creative Cloud Experience Process) is related to Adobe, search their forums for the procedure, probably reinstall that bloated unstable piece of software. (better disable CCX from startup programs) Vcruntime140 is C++, probably missing or corrupted, reinstall Microsoft Visual C++ 2015 - 2022 Redistributable, you can find the installers in ..\DCS World\distr folder, you can also download them from MS. For the sake of it, run a SFC /SCANNOW from CMD.