[SOLVED] Ryzen 9 3950X won't allow over 520MHz on all cores on MSI B550 Gaming Edge Wifi

TheDarkOne198

Distinguished
Jan 9, 2012
184
0
18,710
5
I have had all sorts of issues lately, with multiple reinstalls of Windows 11 and BSODs caused by what I finally realized was Secure boot not letting winload.efi from working. Now that I have it up and running again, I have found that everything is extremely slow and my 3950X is stuck at 0.52 GHz on all cores. I have tried everything from clearing CMOS to updating to the latest BIOS of my MSI Gaming Edge Wifi Mobo. The ONLY thing that helps is, once Windows ever so slowly loads to desktop, I have to open Ryzen Master and disable "prochot" , which lets it go to its set clock speed. If it's not set to "manual" and is on "default" or "Precision Boost Overdrive", it will run at 520 MHz. Is there something in the MSI BIOS that I can set to fix this? I already tried disabling C States and setting it to "Typical Current" under the CBB section, to no avail. It even loads into BIOS extremely slowly and once yesterday,it didn't load properly and partially was in Japanese. Not sure what's up with that. I am hoping it's not a CPU problem and more a BIOS issue but updating the BIOS, which of course flashes a new version, should have fixed that right? Is there anything I can do besides disabling "prochot" every time I reboot? Would maybe flashing BIOS again be advised?
 

Darkbreeze

Retired Mod
I think, if it were me, I'd RMA the motherboard. It might be worth trying a couple of other things first though.

One, I'd make sure the CPU cooler is evenly tightened all the way around with no one spot tighter than any other so the CPU isn't getting "cocked" in the socket, which can cause all sorts of mystery ghost problems AND that it isn't simply overtightened. As well, removing the cooler and CPU and double checking that there are no bent pins at all might not be the worst idea either. Also that there are no loose screws or standoffs in the wrong place between the motherboard and motherboard tray portion of the case.

If all of that is good, I'd try disconnecting all drives except the OS drive, and if the problem still continues might also try using a different drive for the OS, BUT still with all other drives not connected.

If no love there then unless something really goofy is going on with the power supply it almost has to be a board problem.
 

Darkbreeze

Retired Mod
There are only two reasons these low frequency caps happen most of the time, especially on relatively new systems, and that is

1. CPU is "actually screwed". Which is REALLY uncommon.

Or,

2. Motherboard is fracked. And that is NOT uncommon.

I'd suggest that you list your FULL hardware specs, EVERY SINGLE thing you've already tried AND ANY other relevant information that you've not already offered.

I mean, there are other reasons why this can happen, especially if it's a laptop, but it's still not specifically "common" without some other conditions being met.
 

TheDarkOne198

Distinguished
Jan 9, 2012
184
0
18,710
5
There are only two reasons these low frequency caps happen most of the time, especially on relatively new systems, and that is

1. CPU is "actually screwed". Which is REALLY uncommon.

Or,

2. Motherboard is fracked. And that is NOT uncommon.

I'd suggest that you list your FULL hardware specs, EVERY SINGLE thing you've already tried AND ANY other relevant information that you've not already offered.

I mean, there are other reasons why this can happen, especially if it's a laptop, but it's still not specifically "common" without some other conditions being met.
Its a PC. It all started when I changed hardware,getting a new GPU. I made the mistake of forgetting to uninstall the drivers first and got driver signature BSODs. I now realize if I had gone back to my original configuration,it probably have booted but I was too frustrated to think straight so I instead did a clean install of Windows 11. At first it went well but then I got Winload.efi BSODs and sometimes NTOS.Krnl ones,which I now realize was caused by Secureboot being enabled,but that's never been a problem before. After a few more install attempts of windows later,I noticed it being very very slow in everything and this was before disabling secureboot as sometimes it would load windows,albeit slowly and I managed to install some updated only to lead to the winload.efi crashes. I also changed to a different install drive as I was thinking the one I was installing on may have had issues. I also soon noticed that in BIOS,it was saying 520MHz as the frequency even though it was set to defaults of 3500 base. But now it reads what I set it to with Ryzen Master,although it still goes very slow so it must be reading it wrong. My motherboard and CPU are only a year old or so. I never overclocked,other than using PBO and now every time I (slowly) boot windows,I have to make my way into Ryzen Master and disable prochot,which from what I understand disables thermal limits for extreme overclocking. I find it hard to believe installing windows 11 a bunch of times could have messed things up,much less installing a new GPU. I do not understand why its doing this.

As for my specs:
Ryzen 9 3950X w/ NH-D15
MSI B550 Gaming Edge Wifi motherboard with latest BIOS from MSI
32GB G.Skill Aesir
Radeon RX 6750XT (was RX Vega 64 with a RX 580 for mining)
Windows 11 on Samsung 970 Evo Plus 1TB NVME
Cooler Master HAF XB Evo with plenty of fans
Other storage:
1TB WD Blue HDD
2TB WD Blue 2.5" SSD
512GB Plextor 2.5" SSD (was the OS drive before)

I feel its most likely a pure BIOS issue but I flashed it with a new version and it didn't help. The only physical changes made were the new GPU and then multiple attempts to install Windows 11 after that. IS there anything else you need to know,I cannot think of much more but maybe if you can,I can answer.

EDIT:Also,one time BIOS failed to load right and did it partially in Japanese for some weird reason and after I disable prochot,it runs exactly as it used to. Set to 3.98 GHz,which used to be what it more or less boosted all cores to with PBO enabled and voltage is set to auto so stays in the 1.2V range,near as I can tell. Still gets the 20K in Cinebench R15 it always did and only gets up to about 81C while running that test,as it always did. If there is a way to disable prochot in BIOS or have some other work around,it should be okay...I hope.
 
Last edited:

TheDarkOne198

Distinguished
Jan 9, 2012
184
0
18,710
5
I reflashed the BIOS and it didn't help any. If anything,it made things more weird. It loaded partially in Japanese again afterwards and for some reason disabled the drive its supposed to boot from,which threw me off for sure. I had to re-select the 970 Evo Plus again. Then it loaded fine,only at miserably slow speed. Back into Ryzen Master,disable prochot and back to full CPU speed,although that stutter is still there and freamrates are low still in games. I really do not know if its the CPU or motherboard going bad but I can't afford to replace either right now...

and I don't think stuttering is right. Its more like everything freezes just a little and I notices its taking a MUCH longer time to POST
 
Last edited:

Darkbreeze

Retired Mod
I think, if it were me, I'd RMA the motherboard. It might be worth trying a couple of other things first though.

One, I'd make sure the CPU cooler is evenly tightened all the way around with no one spot tighter than any other so the CPU isn't getting "cocked" in the socket, which can cause all sorts of mystery ghost problems AND that it isn't simply overtightened. As well, removing the cooler and CPU and double checking that there are no bent pins at all might not be the worst idea either. Also that there are no loose screws or standoffs in the wrong place between the motherboard and motherboard tray portion of the case.

If all of that is good, I'd try disconnecting all drives except the OS drive, and if the problem still continues might also try using a different drive for the OS, BUT still with all other drives not connected.

If no love there then unless something really goofy is going on with the power supply it almost has to be a board problem.
 

TheDarkOne198

Distinguished
Jan 9, 2012
184
0
18,710
5
I think my cooler is okay. According to HWMonitor,all cores and package and CCDs are in the 40-50 range while idle in windows at 4 GHz and I am very careful to make sure all screws are tight and not overtight and not to bump that giant noctua heatsink so I don't think any pins should be bent. I might try to check for any screws that are loose but it will be hard to do that in my current living situation. Not much space. I am more and more thinking its a board problem too. Taking a long time to POST, sometimes not starting to POST at all (it did that earlier after flashing) and often hanging on VGA. I know drives can cause that so I might try unhooking some of them like you said,starting with that Plextor drive. Its suspect to me and remember,I already relocated the OS install to a different drive.
 
Last edited:

TheDarkOne198

Distinguished
Jan 9, 2012
184
0
18,710
5
I do not think I can RMA the motherboard anymore but I maybe can eventually replace it. I hope it lasts until then.

Edit: I first tried removing that 512GB SSD because I thought it was suspicious but has changed nothing. I was feeling it was probably the board because BIOS is being weird. I probably will try to replace it but I hope everything lasts until then.
 
Last edited:

Darkbreeze

Retired Mod
Motherboards generally have a 3 year warranty. Since the B550 chipset family was released in June of 2020 there is no possible way for you to be outside that window. That board MUST still be under warranty IF you are in the US. For other regions you'll need to contact MSI to find out as it varies from region to region.
 
Reactions: straightawaykid

TheDarkOne198

Distinguished
Jan 9, 2012
184
0
18,710
5
Motherboards generally have a 3 year warranty. Since the B550 chipset family was released in June of 2020 there is no possible way for you to be outside that window. That board MUST still be under warranty IF you are in the US. For other regions you'll need to contact MSI to find out as it varies from region to region.
Is it still under warranty if I never registered it? Every time I tried to register it,it said the model number wasn't valid. I never understood why and when I tried to email their support they never responded. I will try to tear everything apart this afternoon and look for any screws and such. I do know I somewhere lost an M.2 screw and have no idea where it fell to. It is probably under that table the PC is on or something but better check under the board just in case. I already moved the OS from one drive to another so I will make that my opportunity to unhook all the other drives too. I sincerely doubt the cooler and CPU are needing to be removed but I might check anyway and repaste.
 

TheDarkOne198

Distinguished
Jan 9, 2012
184
0
18,710
5
UPDATE: I undid everything,took the motherboard out,during which I noticed there was a standoff that was kinda stuck inside the motherboard and was stubborn to let the board go and it was the one dead center. When I finally (and gently) coaxed it out,I moved it to a different position and made sure the others were fine too. After putting everything back together,I tested each of my SATA drives one at a time in BIOS,which acted perfect each time. No slowdown. Now back into Windows,its running just as it should although I notice a strangely high idle peak core voltage. Its idling at 1.42 roughly but when I do something like run Cinebench R23, it drops down to 1.38 and gets no hotter than 82C and idles at 46C. In BIOS,when I was tried to turn on "Core Boost",it set it to 4.2GHz at 1.4V. I don't think my NH-D15 can handle that so I turned it off. But,as it is right now,unless what I have described is unusual,I think it is back to normal.
 
Last edited:

Darkbreeze

Retired Mod
See, these are the things we FIRST look at when these kind of "ghost" problems start happening. They are not typical. Usually these sorts of issues happen with standoffs, bent pins, or in some cases actual CPU problems which do occasionally occur. I'm glad you got it sorted out and don't feel bad at all, these are the kinds of "learning experiences" that we ALL go through in order to gain some levity when it comes to sorting it all out. If you can't laugh at things, you'll cry. But seriously, glad you got it sorted brother.
 

TheDarkOne198

Distinguished
Jan 9, 2012
184
0
18,710
5
See, these are the things we FIRST look at when these kind of "ghost" problems start happening. They are not typical. Usually these sorts of issues happen with standoffs, bent pins, or in some cases actual CPU problems which do occasionally occur. I'm glad you got it sorted out and don't feel bad at all, these are the kinds of "learning experiences" that we ALL go through in order to gain some levity when it comes to sorting it all out. If you can't laugh at things, you'll cry. But seriously, glad you got it sorted brother.

ANOTHER UPDATE: It just randomly turned itself off and now I get BSoD with an error code 0xc0000428 which apparently has something to do

I know. I may not be out of the woods yet because I was just watching youtube a moment ago and the entire system just turned off. As if it lost power...when I restarted I got a BSoD I have seen before in the past several days, with an error code of 0xc0000428. Something to do with not finding the files needed to boot. I turned SecureBoot back on and it started fine but it was the random power off that worries me. I scheduled a CHKDSK /R as a regular run in Windows says everything is fine but I am worried something might still be wrong.

Could have been my GPU overheating as I just noticed it (mining) and reaching over 90C and over 100C Junction on the stock fan curve and FanControl might have been doing screwy things with it. I set a curve in Radeon Settings and its cooler. We'll see if it happens again.
 
Last edited:

Darkbreeze

Retired Mod
I would HIGHLY recommend that you immediately back up any and all important files, folders, settings, etc., and do a clean install of Windows so we can eliminate that as a source of potential problems. And then, if it immediately continues, address the possibility/probability of it being a drive issue. In fact, back up, then test drives right away. Go from there.
 

TheDarkOne198

Distinguished
Jan 9, 2012
184
0
18,710
5
I have already clean reinstalled Windows on a different drive once. I sadly do not have a means of backing up bu what should I test the drives with? I have already done CHKDSK on all drives without issues.

EDIT: In Windows,within Disks & Volumes,all drives show as healthy as well.
 

TheDarkOne198

Distinguished
Jan 9, 2012
184
0
18,710
5
Hard disk sentinel. Free version.
I installed it. All drives read as healthy,with the only one not showing 100% health being my 970 Evo Plus that is now the system drive (it used to be just for games) and its at 90%. I did a short test on all four of my drives and no problems were found.

Edit: It does warn the temp on the 970 Evo Plus is too high,averaging 58C and maxing out at 62C. It has a heat sink on it though and my fans are always at 100%. It does have a large 6750XT but also a NH-D15 and its two fans right above it. I always felt there is plenty of airflow with my two Silent Wings 3 fans pulling air in. I went with them to force air past all the cables as I have an ATX desktop (HAF XB EVO) case that can be a little cramped with my two GPUs (my old Vega 64 blower card for mining) and their cables that have to go down into the basement right in front of the intake fans.
 
Last edited:

TheDarkOne198

Distinguished
Jan 9, 2012
184
0
18,710
5
Well, I think maybe you're back to it being a motherboard issue. Board was never replaced right?
No it has not been. I cannot access my Amazon account to RMA because of 2FA and not having the proper phone anymore. When it rains,it pours. I might just have to buy a new one on newegg or something...I will hold out with this one until I can't though. Usually, it seems okay. If it gets worse,I will try to buy another.
 

TheDarkOne198

Distinguished
Jan 9, 2012
184
0
18,710
5
Update on the situation. The low-speed CPU issue does still happen sometimes but now the board often does not even POST, with it trying to but only barely flickering the CPU light. Eventually, it might get past it and properly POST and sometimes forcing power off and on again might get it through. I went ahead and ordered a relatively cheap MSI X570-A Pro board as I cannot afford much more. I hope its not a CPU issue as I definitely cannot afford a replacement and I sincerely do not think its a PSU issue. It really feels motherboard,since its having a hard time POSTing...I definitely cannot keep ordering things to troubleshoot. The entire system just froze as I was playing Elden Ring I had to reset,leading to it having a hard time POSTing again. Other times it goes through it without issue.
 
Last edited:

ASK THE COMMUNITY