Akula

BIOS, Nividia drivers and Windows all up to date. Ran a test this morning before updating the Nvidia drivers. Prior to Nvidia update and HT support OFF - Desktop to main menu > 42 seconds Exit to Desktop > 5 seconds After Nvidia update and HT support ON - Desktop to main menu > 70 seconds Exit to Desktop > 40 seconds I don't claim to know what the problem is but there does seem to be a link between Multi/Hyper threading and DCS load times.

I don't think this is a BIOS issue. I think the issue comes from the various updates DCS has made that inadvertently impacted multi-thread support. Something happened because DCS got nothing but praise when they released an update that specifically addressed a significant improvement to Multi-thread support, and they delivered. Since then I think one update was released in 2024 to address a multi-thread bug but nothing after that. We've all been down this road before where one update fixes A, B, C, D but breaks X. It's not the Devs faults as they can't be expected to beta test every single different configuration out there but when it comes to a vital feature like Multi-thread support, it's a feature that needs to be tested after every update on various machines to ensure compatibility/stability. This is free performance that we cannot use due to a bug in the updates that wasn't there before. Obviously I'm not a DCS dev but if I was one I'd get my hands on a 14900k machine and start running some tests. This does not look like an isolated incident.

I would have NEVER told you this in Feb 2024 when I built my rig but TODAY, in Sept 2025 you should be OK. From my chair the problem stemmed from Intel's neglect in setting the Intel defaults for the Motherboard manufacturers and instead of perfecting the Microcode themselves, they left it to the motherboard manufacturers to do it. This ended up with the most base default setting driving up the CPU voltage and heat creating a level of degradation that wasn't normal and would basically kill these CPU's in 6 months to a year if left unaddressed. As it stands, Intel forced all the manufacturers to follow the same template for the microcode so that the advertised and labeled "intel defaults" really were safe parameters set by Intel. All that to say, you should be fine. - Now if you are looking at another CPU and MB combo, I'd also recommend AMD's 9800 X3D.... If I had waited long enough I would have opted for that build.

12600K is what jumped out here as the 14th gen CPU's did have a VERY rocky start. It's over now and the voltages are under control so it's now degrading just as normally as any other CPU. Ran a Cinebench test to see if it is scoring lower than it used to and so far it's holding steady. Score remains unchanged from one BIOS to the other. I checked HAGS and all the other recommendations, all good. Shader cache doesn't seem to help unless you leave it on Driver Default. Curious to see what others on a 14900K are experiencing.

Thanks for the help, SD Quick update although it may not last…. I updated my BIOS and since I can’t run all 4 sticks with XMP enabled and at its rated 6400Mhz, I backed off the memory speed from 5600 CL36 to 5400 CL36 and locked in my voltage to under 1.3, watts at 307 Still running my 14900k at 5.5Ghz tested it twice and both times I was able to reach the desktop on safe mode to make my changes. If this changes or “breaks” I’ll report back. conclusion: it appears memory timings and memory settings may affect windows ability to boot into safe mode but it’s just a guess. So far all things point to BIOS update since changes in the older BIOS wasn’t helping but we will see. So far so good

Quick update: BIOS, Nvidia and Intel Arc update hit me last week so I just updated *everything* and tested again with hyperthread support enabled. Double clicked the regular BIN executable (not the MT executable) From desk top to DCS main menu with HT enabled - 2:03, as in two minutes, three seconds From desk top to DCS main menu with HT DISabled - :43, as in forty three seconds This isn’t a user issue. I hope ED looks into this because it looks like hyperthread support is broken in DCS. For better transparency I did notice that the max CPU temp to bring up the DCS main menu was 73 with HT enabled and 58.5 with it disabled. Seems par for the course. - Confirmed all cores UNparked.

Thanks for this, Im currently caught in an update storm. Nvidia, new MS update to address the corrupting Western Digital M.2 SSD drives as well as a BIOS update. I’m going to try updating EVERYTHING as it all just hit me at the same time. I doubt this will fix anything but if I’m wrong I’ll happily take it lol. I’ll update soon. Wondering if anyone else has the same MB and CPU on a 3090 setup. Also, my memory was stable as hell at 5400 vs 5600Mhz. As I’m running 4 sticks on a dual channel board (yes I know this is stressing the memory controller and yes I know I should be using 2 sticks instead of 4). The memory is corsair vengeance DDR5 rated to 6400 in XMP but XMP is disabled to run the 4 sticks. Stay tuned

Thanks for the update. I should have posted a correction a long time ago. I currently running 576.52 and the problem came back so I'll have to troubleshoot further to see hat the problem is. If you don't mind, could you try it again? I'm just wondering if I was able to get into safe mode because the reversion wasn't followed by a restart prior to trying to get into safe mode.

I don’t think so….. I currently have hyperthreading disabled again and DCS runs incredibly well. fast load times, preload radius loads very quickly vs seeing a slow transition of water to ground terrain then buildings and so on. To be fair I will add that I did change my memory clock from 5400 to 5600 using “memory try it” in my BIOS. I suck at manually tweaking my memory so I had to use MSI’s feature. That little adjustment sped up my load times, BUT - if I re-enable hyperthreading it basically kills any hope of flying without some sort of lock up. I really do think that there is an issue here and I hope ED really does take another look at hyperthread support because it is like night and day when it’s disabled vs enabled. In all my experiences hyperthread enabled ruins the game vs hyperthread disabled fixes every long load and stuttering issue. I will do a little research into tweaking my memory manually and see if it’s a memory issue conflicting with the CPU but it doubt it’s a factor since I never had to do anything like this after the update that was released almost a year ago that directly addressed full and hyperthread support. That update was amazing. Since then it has been touch and go with hyperthread support. one update fixes it and another ruins it and so on. Curious to know how others are doing with this.

THIS IS WHAT I USED TO DO BUT NO LONGER WORKS In a mission with three bandits and me with 6 aim120c I can usually target all three jets and fire on them consecutively. This is how I used to do it select TWS mode set to AUTO let the lead jet get automatically locked in TWS and let the track file populate. At 80 miles i back out of radar and select “ANZEL” and the “AUTO” Back to radar select EXP to expand the view. select RAID from there I can press “S” to toggle through the targets in TWS and fire at will. That method no longer works. When I set it all up similarly, pressing “S” doesn’t toggle to the other contacts. Just stays locked on the lead jet. with the way it is now I manually have to fire on the lead, break lock, lock up no. 2 and repeat for the third jet. I’m pretty sure this is broken unless the procedure has changed.

Disable hyper threading and try it again. ED broke hyperthreading support (again) a few updates ago.

Admins, apologies if this is in the wrong section so please feel free to move it if you need to. _____________________ if you are running an RTX 3090 and on the latest Nvidia driver (580.97 as of August 16, 2025) and CANNOT get into safe mode, this might help you. problem: cannot get into safe mode what is happening: - In windows 11. - press start > select power icon to restart > hover mouse over restart > hold down left-shift and click restart to bring up troubleshooting window - trouble shooting window comes up, no problem so far. - Select Troubleshoot > select Advanced Options > select Startup Settings > select Restart. No problem so far……. BUT! The system boots to a black screen and there is no blue screen that allows you to select option 4 or 5 to boot into safe mode. Here is how I *thought* I fixed it. - pressed Enter > Rebooted back into windows 11 - Used “Direct Driver Uninstaller” to uninstall the latest Nvidia driver (580.97) - tried booting into safe mode again and this time it worked. I got into safe mode and used DDU again as a failsafe to remove every single trace of that POS driver. - Now back to windows 11 - installed driver version 576.52 - Tested again and I was able to boot into safe mode. fixed! Or so I thought. A week later the problem returned but now it's going black screen, timeout and automatically shuts down. Nvidia seems to have crapped the bed with their latest drivers. I don’t care what anyone says, to me it looks like they are so into trying to revive the 50 series cards that their latest drivers are absolute shiiite for 3090 users. In light of the most recent findings. It could be an Nvidia issue or it could be a windows issue or a MB BIOS issue. So far, Google has not helped. Still looking into it. anyways, hope this helps someone.

For poops and giggles I tried the shader cache fix….. I’d don’t really help but at least I wasn’t getting insane stuttering in the game. for now I’m just going to leave hyperthreading disabled as that seems to be the only tried and true fix for all the symptoms described in this thread.

Just confirmed HAGS has been off since I last disabled it a year ago but I guess windows re-enabled “Game mode” after a few updates. I just disabled it again. Might be re-enabled when I play steam games but I’ll test that later. For now it’s all disabled thanks for your help.

Thanks Mainstay. Here is my latest log file as requested. - I unparked my cores as soon as I installed Windows. Just checked again and they remain unparked - I also delete my FXO and metashader folders from time to time - As well, main DCS folder along with my DCS saved game and Simapp Pro Folder are excluded. The flagged solution is one i have yet to try so I'll give that a shot as well. === Log opened UTC 2025-07-31 19:51:59 2025-07-31 19:51:59.570 WARNING EDCORE (Main): hypervisor is active 2025-07-31 19:52:01.033 INFO EDCORE (Main): Platform Timer Resolution: 15.83ms 2025-07-31 19:52:01.522 INFO EDCORE (Main): Created boot pool: n:24 2025-07-31 19:52:01.522 INFO APP (Main): Command line: "D:\Program Files\Eagle Dynamics\DCS World\bin/DCS.exe" 2025-07-31 19:52:01.522 INFO APP (Main): DCS/2.9.18.12899 (x86_64; MT; Windows NT 10.0.26100) 2025-07-31 19:52:01.522 INFO APP (Main): Application revision: 252899 2025-07-31 19:52:01.522 INFO APP (Main): Renderer revision: 26305 2025-07-31 19:52:01.522 INFO APP (Main): Terrain revision: 26363 2025-07-31 19:52:01.522 INFO APP (Main): Build number: 365 2025-07-31 19:52:01.522 INFO APP (Main): CPU cores: 24, threads: 24, System RAM: 130766 MB, Pagefile: 8192 MB 2025-07-31 19:52:01.522 INFO APP (Main): CPU info: system affinity mask: 111111111111111111111111 process affinity mask: 111111111111111111111111 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} logical cores with efficiency class 0: {8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23} CPU cores have different performance classes: [0-1] logical cores with performance class 1: {0, 1, 2, 3, 4, 5, 6, 7} logical cores with performance class 0: {8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23} common cores: {4, 5, 6, 7} render cores: {0, 1, 2, 3} IO cores: {8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23} unavailable cores: {} pause10: 0.030879 us (std dev: 0.021198) pause10: 973 cycles (std dev: 652.9) Cheers and thanks