Numbafieve

hard to say for sure but it could have been so tight on either one side or in general that it was warping or off center. Tighter is not always better for many things involving screws lol. Thats a more acceptable yet high idle temp.

https://www.cgdirector.com/msi-ez-debug-leds-troubleshooting/ Here is something i found, you may not have Q-codes but MSI has something half as good. If you go through it you may find something. If not def report back here boss.

Im still thinking a bad mount but that also wouldnt explain the expected hot temp under load (for the most part)...

if its an ASUS mobo then they usually have a way to do it via a USB drive written to from another computer, If its not an ASUS mobo its more likely you would need a CPU that is already compatable (13th gen). But before we do anything that extreme do an idiot check and walk yourself through all the stuff you plugged in and installed when building it. Im assuming you build this computer yourself from off the shelf parts? Also if you could tell us exactly what parts you have that would be super cool cuz we could help more if we knew what we were working with.

Awesome straight from ze manual. I looked through it but couldnt find anything specific to my case. I did see something about possibly needing NV link to increase memory from somewhere else. Guess thats the way I will have to go. Probably will try it without first but I have to do what I have to do I guess.

they did say there was a download in the background so its not like the CPU is totally parked rn. My CPU sits at 4.8 all the time just due to any background process un parking it.

Correct, even a basic cooler should be able to keep it cooler than that at idle. What cooler and cpu is this? It sounds like probably just a bad mount tbh. If you dont have additional thermal paste it may not be an issue yet either, usually you can just re mount and it will be fine.

Im guessing there is no Q-Code readout on your motherboard? That would be the most helpful at this point really. If not then from what I found 12th gen may need a more up to date BIOS to function on that newer chipset.

I am currently working on a huge render in blender. I have found i need about 120GB of system memory and about 50GB of VRAM to do this with no page files or memory swap into system (which i dont have enough of anyway). Right now i have a 5950x (OC to high hell, 4.8GHz all core) 64GB of CL14 3777 DDR4 3090 + 1050ti to run extra monitors all water cooled and with a good PSU for now The problem is that I am obviously running out of both types of RAM, Blender even tells me this. I have a 128GB kit of DDR4 showing up later today and am currently looking at buying a slightly broken matching 3090 that I can fix. I have been looking but cant find any definitive answer as to if having 2x3090s will stack the VRAM or not. I know in the old school SLI days it did not but I was under the impression that most render engines like Blender that was not the case and the GPU memory could actually stack. If so ill probably get 2 more 3090s and a 4090 for this system and then build a render farm for future use with the 4x3090s and 512GB of RAM. If anybody has any experience or can find some bit of info about blender I wasnt able to that would be fantastic. I cant even attempt this render until I get some kind of fix for this issue, and obviously I cant get paid for this project until I do actually get it rendered.

I enjoy CPU overclocking quite a bit and was having fun with my 5950x tonight. I was drinking heavily since I just had an argument with my girlfriend and was screwing around with my computer as I usually do when im alone. I know this would not fly with a true overclocking rank but I know I didn't cheat or anything so i'm super happy with this score of 13020 in Cinebench R20 that I got with my 5950x. None of my friends care at all about overclocking so I decided to put it here where it might actually be appreciated a little bit. I don't know why Cinebench isn't showing the true clock speed but it was 4820MHz all core with a super slow RAM speed of 1600MHz CAS 20 TRaS 30. I was not expecting to get this score tonight.

Yes afyer a repaste they are much better. No more thermal problems. What is interesting is that my desktop lasted close to 3 years with no repaste and it still didnt have any problems.

Back in 2017 ish I purchased an alienware 17 r4 for college. It has a GTX1070 and the i7 7820hk which I overclocked to 4.7GHz. It was a champ at everything I wanted it to do and when I switched peofiles to my underclock of 1GHz it would go for a long time on battery for my classes. Obviously it made a ton of heat and so I fixed the cooling on it. Both the cpu and gpu got LM and all of the thermal pads were replaced with fujipoly ones. The chipset even got its own little heatsink feom a raspberry pi. At my 1GHz setting it never saw above 40C, at stock 2.7-3.something it never went above 65c and at my overclock of 4.7 it never hit 90c. The cooler was doing a really good job thanks to iunlock. Also a lot of time tuning the undervolt. It has been 2 years since then and I dont really game or benchmark it anymore. So when I did finally launch a game again I noticed it would hit 99c instantly. I opened it up and checked the LM. The LM had dried up, mostly on the cpu (smaller of the 2 dies) while the gpu was still ok. The gpu never really got all that hot, even stock. The lighter colored spot is actually a solid chunk. Here is the heatsink side and it didnt fare any better obviously. It took the most work to clean up since the LM soaks into the copper while it pretty much just wipes off the dies, usually... I had to use a razor blade to get the dried LM off the dies as well as do the same for the heatsink. Afterward I polished the heatsink since it was super scratched up. here is a little closer shot of the heatsink. My phone camera kind of sucks sorry. I lost the pictures of the after but the LM left a stain on the heatsink and the dies cleaned of nicely. I didnt do LM again now that this laptop is just a fancy youtube machine. The only reason I am posting this is to report my experience with LM not to criticise it. If there are any questions I will try to reply today but it may be pretty late before I can check.

yep tried to reset CMOS already and it didnt help. I forget what the place was called that i sent it to though. They didnt even send me a picture or tell me what was "wrong" with it which is why im pretty sure they did nothing. I just dont know if i should start by replacing the MB or CPU first.

I just looked at the link you gave and yes the 3rd one is actually a 16x. The post I was looking at was wrong on the PCIe sizes. After reading some articles it also looks like the card is fine and this is 100% windows fault. have you tried this yet?

3rd is a 16x size but only wired for 8x. the second to bottom and very bottom have a 16x "link". I doubt this is the problem but it could be.