BitMaster

B-Die slowly looses grip beyond 4000MHz and that's were Micron E-Die takes over with exceptional high frequency stability, tho at looser timings but the overall gain in MHz compensates for the looser timings so latency more or less stays the same while bandwidth increases. Not relevant for any modern AMD CPU as we all know, somewhere shortly beyond 3600-3800 most/all IMC start acting up. more or less what Buildzoid says

The fan ! If you had not gamed but only read the news in a browser it could have taken forever for this broken fan to make a difference. Glad you fixed it !

#1 No Mercy from any Admin if you don't have a Backup ! #2 Make a Backup while you can !

Do the math, if you really run it for a longer time, self-hosting pays. The Gigabit line has an UPLOAD and a DOWNLOAD. The server needs as much upload as it needs download. Depending on the UL speed it may work with 20, 30, or 60 players. Download is plenty. My current cable Gigabit has 50Mbit Upload only, my coming FTTH Gigabit line has 400Mbit Upload...that's another level of pushing data out.

Do not overclock your card in DCS unless you have a really good cooling..and you do not have that with air. Try any stability test w/o ANY overclock as a baseline. Maybe even try out older or newer NV drivers. Often your scenario originates in driver issues tbh. Use DDU to uninstall your present NV driver, do it properly, with the reboots and all that as described and then install another version. But don't OC when you search for sudden reboots or BSOD's, that's a bad start.

It's likely not the CPU that's causing an issue. Games usually do not tax all cores to the max like any good stress tool does. Most test suites do not test CPU and GPU, or just the GPU. In games, if the CPU ain't too weak your GPU will always try to run at it's fullest potential unless otherwise told to limit fps. That's where the most constant power draw and heat comes from. Modern CPU almost fall asleep while gaming on 1-2 cores and idling on all others, the GPU no so.

Got mine from amazon back in 2015, try there.

create a new user account, start DCS from within there and check if the error carries over to the new user. If it does, well...I haver no answer to this but if it doesn't it is likely a corrupt setting in your DCS Home Folder which you could delete/fix and have it solved.

40 Players need a lot of bandwidth too, not only proper hardware. Are you aware of that ? Do the math, if it pays to run your own rig. Electricity ain't for free so is the rented server. You always pay for a rented server, if you use it or not, it's a monthly thing, reoccurring. With your own rig, you can fire it up when you need it and turn it off when you don't and save some money. Contrary, if you want to let it run 24/7 it can quickly become very expensive to run a server 24/7, I run a few and I tell you it ain't cheap, it's big bucks at the end of the year. There is no need to get the fastest multi-core CPU with lots of RAM for a DCS server. I would either pick an Intel with iGPU ( low core count ) and DDR4 or if money allows get a new Ryzen 7000 with iGPU but it comes at a much higher price. Avoid anything that doesn't read INTEL or BROADCOM on the network card, in server land you dont want to read REALTEK ( fighting one now as I type ), just a tip from me. The big obstacle will be a proper connection with no fluctuations, jitter, ping losses etc.. and a router that can handle that many connections without dropping packets. Maybe a standard home user router is not the right device then, it depends on it's switching speed aka CPU power. Most home user routers are not really FAST and 40-80 conns is not nothing.

So it seems I was right with a faulty fan. Now, follow Skatezilla's advice and get a BIGGER Powersupply. You never cut corners on a PSU, you will suffer if you do out of my experience. edit* ....and yes, the Mobo VRM is borderline critical. You don't pair a V12-800BHP engine with a Fiat 500 Chassis, same applies for your CPU and Board. Epic mismatch

It could be a simple PSU fan failing, slowly overheating your PSU. Been there with my 1st AXi-1200 which had a defect off the shelf. It ran flawless in Desktop use, no real load & heat, but shut down the PC after some time gaming. Verify your PSU fan is working and that the PSU case is not hot by the touch after it shut down by itself. If it is, you found the bad part.

You can’t fly properly without rudder pedals. crosswinds, slip approach, aiming correction, wing drop trim, etc etc “Fly the Ball“

But that ain't B-Die from all I can see

Me neither, but I still run a 1080ti. Might be different with other SKUs

There is one counter, one employee doing the whole show. If only 2, 3 or 4 guests are in it's all OK. Everybody gets their Soda and Food in time...just when Joe and his Football Team come through the door and every table and seat is full it would be SO MUCH BETTER to have 3 Girls serving the crowd, 2 guys doing the drinks and such and a great cook that has nothing to do but cooking. One day we will get there, be patient.

To say it short: HIGH MHz & low latency is the goal

Try DP first.

Actually, I feel even more so when I encounter one of the many shortcomings that Linux Desktops have. - Can't install any of the recent ( ~2 years ) Debian/Ubuntu/Mint unless in compatibility mode, it just won't digest my 1080ti until I install the NV Meta driver. Not an issue for me, I can run it headless and work on shell, for many the game is over at that point before it even started. - Numerous Sound issues with multiple speakers/headphones connected.No proper Realtekcodec implementation ( also due to lack of source code to be fair ) and no real support for higher end DAC's. They just dont get sound sorted properly, multi-in multi-out doesn't work. No sound on speakers unless I unplug my headphones. At least, they got it working in 22.04. A very annoying bug for any gaming PC and I have this on all recent PC's with "normal" onboard sound found on every midrange board - ZFS on 22.04 can be a mess. Prepare to clean your boot drive manually or you can't new kernels anymore. Not a ZFS bug, a nasty implementation bug that you cant solve when formatted. Real nice ! zfs list -t snap -r -o name,used,refer,creation bpool sudo zfs destroy bpool/BOOT/ubuntu_zu4sdi@autozsys_0n2yi4 LOVE IT ! - I could go on.. I like Linux and I use it everyday, Desktop and Servers, but to say "use it because Redmond treats you as a Beta Tester and we don't" is not what reality looks like. I would pick 11 unless there is a real issue that you cannot solve but they are rare. I run many machines on 10 and 11, they both work great but 11 is nicer to work with if you get used to it and imho it seems to run smoother.

It's many stupid things coming together, LoL, but Igor's photo's and explanation are the best so far for what has happened to the 4in1 Adapters. Still, the 30x connection leave a bad taste, even if they are not foremost a fire hazard.

Oh well, than it's up to Nvidia to fix this. Still, I doubt that another Adapter or ATX3.0-PSU cable will be completely safe. The way the Pins are arranged cries for such things to happen. BTW, the 30x in&out applies to all, not only the adapters and it actually means or refers to all parts involved, incl. the GPU socket. The high load also not jumped onto that cable 1 inch before passing the socket, it came on this wire straight from the PSU. I would double and tripple check my PSU Pins as they also have been running the exact same load on those wires that melted. etc etc.. OK, I am gonna stop here, ain't gonna buy one anyway. EDIT: Igor seem's to have found the real problem, which is actually not the Pins ( I still dislike them ) but a bridge plate onto which the cables are soldered for load distribution.

Afaik the issue occured also on MSI, Gigabyte and Asus boards and again, afaik, those ship their own Adapters. You can't have a female that isn't closed all around, as soon as the male pin wiggles or puts stress on it, or 30x in & out, the whole thing becomes so wide that there is no more full contact given, and that, I am afraid, comes with all those adapters as that kind of male/female is what they use and I think they can only use those. So it's only more or less likely to occur from manufacturer to manufacturer but the error is in the design chosen. The design is ok, for much lower ampere loads. I used to fly high powered electric R/C planes, most where roughly around 12v, give or take. When you pull up to 250w, G2 2mm Gold Contact is ok, 250-500w it's G 2.5 and G4 for higher loads up to 200amp and watts up to at least 13kW is what I have seen working with G4 Gold Contacts. Those not-closed-loop female pins are the culprit imho. Anyway, the damage is done, Nvidia will sell less just because of that and AMD might get some more customers just because of that. What a dumb decision they took.

I think the problem is down to how the female part of the connector is made, not saying the male couldn't be better too. It's unbelievable no one saw this problem before it was too late. How can you make a female socket that is made like that for such a use case??? A solid hollow female that you cannot pry open and a better, solid male. 100watts through one of those pairs.... freaking suicide. When you do electric motors, you roughly know how much gauge you need for cont. 100w and what the plug should look like. More like this, with feathers on the male and solid female: This is a really bad show for all involved. IT SHOWS HOW MUCH THEY CARE but want your money !

Yes, you do, the ONE button at the VERY LEFT BOTTOM of your screen I am located in Germany. Try if you get it done, if not, there is Teamviewer. That is way better than any phone call can ever be. Send me a PM if you need remote assistance to get that file sorted. You will need to download Teamviewer, install it and send me your ID and PWD via PM, NOT IN THIS PUBLIC CHAT !!!!

Easy, right-click on Start button and chose Command Prompt or Terminal as Administrator. you do not need to be in c:\windows\syswow64 to execute the command. Open the terminal and execute the command, that's it.