Installed a new CPU (Ryzen 7 5800X3D) and now at random times the PC restarts.
Posted by buckeyebrownies13@reddit | buildapc | View on Reddit | 107 comments
Replaced my old cpu with the new one. Before doing so I updated the mobo (B450M/ac) bios to 3.10 as that would support the new chip. Everything started up fine. Ran a round of Tarkov just to see the results, temps seemed fine and have zero issues. Then later in the day the PC just started randomly restarting. Seemed to happen pretty often when I was download BF6 on steam(idk if that helps). I’m kinda new to all this so any help would be awesome!
mj34hig44@reddit
Fresh Windows installation?
sorry_but@reddit
I've done a complete swap of most components and not done this. For a simple CPU swap this is definitely not needed.
mj34hig44@reddit
"definitely not needed" is definitely not true. As I said, sometimes you can get away with it but it's far from ideal. If you're going from maybe a R 5 2600 to an R 5 2600X you're probably fine but if you're going from an R 5 2600 to say... an R 9 3900X that would be foolish.
Not the best practice and when you have issues like the OP, clean install is the first step.
sorry_but@reddit
Been building for 25+ years. First step in troubleshooting is not a reformat/reinstall. First step would be uninstall old hardware and chipset drivers, etc. Then run a live Linux distro. Then start swapping components. A fresh OS install is rarely needed. There is no magic to porting it to new hardware - just ensuring the correct drivers are being used which does not require a fresh install unless there's something funky with the removal of the old drivers.
mj34hig44@reddit
Where does the OP say he's running Linux? I've been building since Win 98, 2000, first CPU an AMD Athlon 550 on a BCM board. Windows until maybe Win 7 was always reinstall OS when changing most hardware especially CPU.
As stated above, you may get away with it, but if you're having issues the first rule of troubleshooting is to eliminate variables. An old OS running older hardware is a variable.
IStillHaveFaith@reddit
I built my PC a month ago with the same CPU and had this same problem. Could be a lot of things like PSU, but for me I saw a reddit post saying there was something wrong with a CPU core.
I exchanged the CPU for a new one and it now works without a problem.
buckeyebrownies13@reddit (OP)
So far it’s passed all the stress tests
erikvanendert@reddit
Which stresstests did you run? Im having same issue with same cpu.
buckeyebrownies13@reddit (OP)
I used cinebench. Mine passed single and multi core. Turns out my issue was motherboard
little_lamplight3r@reddit
If nothing else helps, turn off memory integrity (it's under core isolation) in windows security. It uses Hyper-V that Ryzen 5xxx struggles with for some reason.
I spent 3 years trying to fix random reboots. I changed my PSU, updated the BIOS, removed the GPU driver with DDU, and tried countless other steps. The only thing that helped was turning off memory integrity.
buckeyebrownies13@reddit (OP)
Thanks! I will give that a try
MustafiArabi@reddit
i say PSU slowly dying.
I also have the 5800X3D and those symtomps sound like PSU issue. OR Ram but i would say PSU
jhaluska@reddit
Random crashes can be a bunch of things, but PSU is definitely on the list. They're the most annoying to diagnose which is why experienced builders don't skimp on them.
quarterbreed@reddit
Tell me about it....Took me over a year to figure out mine. It would randomly reboot in some games and not others. Turned out to be a bios upgrade for my GPU. What a headache that was.
erikvanendert@reddit
GPUs have bios upgrades? My pc also randomly crashes and havent found cause yet. 5800x3d rtx3080.
quarterbreed@reddit
My 6800xt did. It fixed my issue. I bought new ram and psu in the process that never fixed it but the bios update did
jhaluska@reddit
If my PC is stable, I'm a lot less inclined to upgrade. I'd rather deal with low FPS than zero FPS from crashes.
Doza93@reddit
Brother I just re-learned this lesson the hard way after a recent GPU "upgrade".. I just wanted marginally better graphics, but it wasn't worth the headache tbh
jhaluska@reddit
It's a tough lesson. We often recommend upgrading like it's risk free, but it's not.
autodidacticasaurus@reddit
If it ain't broke, don't fix it.
Chaoselement007@reddit
… but if it is…
buckeyebrownies13@reddit (OP)
Ok, it could be the PSU, it’s at least 5 years old
jhaluska@reddit
Does the new CPU require more watts than the old CPU? You can try undervolting to decrease the power consumption which could actually help stability.
bobsbitchtitz@reddit
It's probably the PSU, you can check your parts on pc parts picker and see what the expected draw to the PSU against the wattage your PSU is rated for. A real energy test will tell you better.
MustafiArabi@reddit
What Brand and Model do you have and whats the Wattage. And 5 years isnt alot for a Quality PSU
KeyCold7216@reddit
Could also be that a few of the mobo pins are fucked. I was having pretty much the same symptoms as OP. Turns out a few of the pins in the cpu socket somehow started to corrode.
MustafiArabi@reddit
ah yes. AM4 has Pins on the Mobo. Im guessing by the lack of education, knowledge and intellect you are an Intel buyer?
withoutapaddle@reddit
Wanna hear something embarrassing? I had a similar problem, and that's how I found out for the first time that cases had a PSU filter on the bottom. Mine was absolutely matted with dust, like a 1/2" thick carpet of it. PSU was overheating under high amp draw when gaming.
CatsOP@reddit
erikvanendert@reddit
How does one check psu health?
Entire_Device9048@reddit
Why are you only on BIOS 3.10? Updating to the latest would be the first step. BIOS 3.10 on most ASRock B450M/ac boards corresponds to AGESA 1.2.0.3 but the 5800X3D requires at least AGESA 1.2.0.6b or newer (sometimes labeled BIOS 4.00 or later).
If that doesn’t fix your problem it’s almost certainly the VRM on your B450. The B450M/ac has weak VRMs, and some boards throttle or brown-out under transient loads, leading to restarts. Use HWInfo64 and check CPU VRM Temperature and 12 V rail stability. If temps exceed 90 °C or the 12 V rail sags below ~11.6 V during load, suspect VRM or PSU instability.
buckeyebrownies13@reddit (OP)
I used 3.10 as it was the latest stable update. Should I download a beta update?
Entire_Device9048@reddit
If your 3.10 release notes specifically show AGESA Combo V2 1.2.0.6b or higher, then you’re already on a recent firmware that officially supports the 5800X3D. If not, then you should update to one that is.
DWeh_@reddit
Had that Same issue, for me it was the RAM, too high speeds. Had to make the CL higher
buckeyebrownies13@reddit (OP)
What’s the fix there? What do I need to do?
DWeh_@reddit
Like I Said, you need to make the CL Timings of the RAM Higher, or give the RAM more Volltage, upto 1,4V is okay I think
kingofneverland@reddit
Where is that setting? BIOS?
DWeh_@reddit
Yes, xmp/expo, however your Motherboard calls it
buckeyebrownies13@reddit (OP)
Tried this, no luck so far
DWeh_@reddit
Could also bei a faulty USB device, which shortens the Motherboard, but would be weird after a CPU swap
buckeyebrownies13@reddit (OP)
Just a bad device plugged in via USB?
zDexterity@reddit
if u undervolted it, then don't.
buckeyebrownies13@reddit (OP)
How do I check?
stickyjam@reddit
do you have a undervolt set?
My msi motherboard is stable at -10, and -20 but is only 99% stable at -30.
buckeyebrownies13@reddit (OP)
Honestly not sure. How would I check?
stickyjam@reddit
Probably easiest way would be to clear settings to default if not done already , otherwise uploads pics of all the bios pages
buckeyebrownies13@reddit (OP)
What page specifically? Not sure where to go for this
NintendoJesus@reddit
When you've exhausted all suggestions here. Find your 'Global C-state' setting in the bios and change it. If it's enabled, disable it, and vice versa.
buckeyebrownies13@reddit (OP)
What does that do?
NintendoJesus@reddit
Basically nothing. But I had a similar problem as you. After many hours of google-fu, I came across a post that spoke about something to do with AM4 boards and 5800+ cpu's. I thought my motherboard was bad, or my cpu was bad, nothing else worked, but disabling c-state did. Haven't had a problem since. It probably won't help you because it seemed to be a rare problem when I read about it, but it might. And you can just change it back anyway. Good luck!
buckeyebrownies13@reddit (OP)
Thanks homie!
NintendoJesus@reddit
Np, hope it works for you.
buckeyebrownies13@reddit (OP)
I had so much fun hope lol. It didn’t help me, issue still happens. Thanks though!
daxtonanderson@reddit
Most likely XMP/EXPO likely turned itself off during the BIOS update process and now the 2400mhz speed is unstable on your kit
Look in your BIOS settings for XMP/EXPO profiles, if you can't find them google your motherboards model number and XMP
buckeyebrownies13@reddit (OP)
Tried this, no luck. Thanks for the help though!
Hive-Mind4085@reddit
You said your temps are fine but what are they?
Did you change anything in the bios after the upgrade?
buckeyebrownies13@reddit (OP)
They were running around 53-55C. I can’t remember what they were during Tarkov. What do you mean after the bios upgrade? All I did was do the flash thing and then restart the computer.
Hive-Mind4085@reddit
Bios flashing resets it back to defaults so checking you didn't change the memory timings or something.
You could go into event viewer in the OS and check the errors and critical logs if it points to anything specific causing a restart. Normally a out right crash / restart is something hardware related like overheating or the PSU though if nothing has changed in the bios / no over clocking etc.
buckeyebrownies13@reddit (OP)
So it may have been my ram timing? It was set at ddr4 2660 but my ram is ddr4 3200.
withoutapaddle@reddit
Turn on XMP in the bios to set your RAM back to the 3200mhz overclock it is rated for.
Potentially your RAM sticks are unstable at 2660, but must be OK at 3200, since the were probably running at 3200 this whole time and the bios update reverted the XMP profile to off.
buckeyebrownies13@reddit (OP)
Tried this with no luck.
scalyblue@reddit
On amd and it would be docp
buckeyebrownies13@reddit (OP)
Ok so get into the bios and check that my memory settings are inline with the ram I have in? Sorry if these are dumb questions. I am new to all this. I had looked into event viewer but nothing really stood out to me.
I have not done any over clocking and to my knowledge the power/voltage is similar to the chip I had in (ryzen 7 3700x)
Hive-Mind4085@reddit
Also just to double check, when you replaced the CPU you did put on new thermal paste right?
buckeyebrownies13@reddit (OP)
Yes I did!
buckeyebrownies13@reddit (OP)
Also, thank you for the response!
gudboj@reddit
My guess would be psu as well. I had similar issues, fine temps, random restarts under load, no luck trying the usual ballet such as reseating ram and gpu, new thermal paste., bios and so on.... new PSU solved it.
kingofneverland@reddit
Does screen go black and pc not answering but still running? Because I am experiencing that. (My cpu is 5800x tho). I tried everything: reinstalled drivers windows etc. Tried different ram sticks. Bought a new gpu from amd (I had an old nvidia). But still persisted so I sent it back. After installing my old gpu, I also changed my ram timings and it seems like the issue is gone. I will try to bring my ram timing up again and if it repeats I will try changing my psu.
buckeyebrownies13@reddit (OP)
So your saying to slow down my ram timing?
kingofneverland@reddit
It seems to work for me now. I am at work now so I will change them and try again. If it repeats it will at least make me decide it is the problem and leave them like that. I will build another rig in a few months anyway.
UserUserDontGetOld@reddit
First I'd do is running a stress test (Furmark for CPU & GPU) to maximize power consumption, meantime monitoring temperatures or both CPU and mobo (prefably using pyrometer to check real VRM temperature).
Suspects in order: 1. CPU cooler is installed wrong thus not working at its best. 2. Mobo VRM overheats. 3. PSU is not powerful enough. 4. RAM issues.
If everything turned out to be ok, but problem still persisted, I'd first disassemble the PC and assemble it again, that also helps sometimes.
MadMax4073@reddit
I've read your comments OP and you said your event log error was WHEA error. If you you are sure its not the RAM - RETURN the cpu. This is old and kinda rare problem that AMD were keep being quiet about. I've had it with my 5600x back in the day and my friend at the pc parts store didn't believe me. We ran so many tests that we could trigger it at any point we like in Cyberpunk 2077. It happens during low loads, it can also happen when watching youtube or do some other low demad tasks in desktop. But its easier to trigger in game where you find where the cpu boosts and then drops the clock but it can't boost back up properly and throws a whea followed by system crash. We returned the CPU for RMA and the problem was gone when the new one arrived. Its so rare that my friend who runs pc parts store had less than 10 cpus with this problem during the whole 5000 series life span. It was the 5600x cpus with just one 5800x3d that got the WHEA problem. Also there was overclock.net thread about this that went for 400 pages or something. Eventually people managed to mitigate the problem with running the cpu at lower frequency or the ram running at stock non xmp speeds. But in my opinion this is not a solution. You paid for something and you are forced to use a lesser variant of it.. So yeah run some RAM tests but in the end if the CPU throws WHEAs left and right its better to RMA it. It's not the PSU. If I didn't experienced this problem myself and went through it I also would have suggested either RAM or PSU but...
Spirit117@reddit
Funny you mention this, i have an older MSI B450 motherboard as well with a 5800x Non 3D and whatever bios was available at the time of install to support 5800x since i came from a 3600x and i have exactly the same issue.
PC will just randomly shut down with a black screen when idling or playing low GPU intensity games. If i play that puts load on GPU its fine, can run all day, but if im working on spreadsheets or something itll shut down sometimes multiple times a day.
I already replaced the power supply with a Corsair RMX 850W gold a year ago.
I dont have a fix for you unfortunately, my fix was to just replace the entire thing with a new 9800x3d and motherboard, since im convinced the issue is something related to B450 boards and Ryzen 5000.
Keep in mind AMD originally never wanted to support Ryzen 5000 on B450 at all, but people bitched and so they started offering beta bioses and im not convinced those bioses are properly stable.
buckeyebrownies13@reddit (OP)
Thank you for the reply. That sounds brutal, and that is what I fear is happening.
Spirit117@reddit
Just find a gpu intensive game that doesnt kick after AFK (like helldivers2 for ex) and let it run for awhile and if your pc runs and runs and doesnt shut down, but shuts down while idling or playing light load game then yea, its the same issue.
I havent had any issues playing BF6 today because that game puts load on my GPU.
rip-droptire@reddit
Random restarts is such a difficult issue to diagnose.
Last time I dealt with it, reverting from Windows 11 to Windows 10 fixed it. The time before that, reseating the GPU resolved it. And before that, it was a new mobo.
So... literally any component can cause it. lol.
ToxinFoxen@reddit
It's strange that you haven't considered whether the new CPU is drawing too much power for the PSU to provide.
Does it have the same power draw as the old CPU or something?
buckeyebrownies13@reddit (OP)
It does, yeah
ButchyBanana@reddit
Had almost the same issue after upgrading my CPU (3200G -> 5800X) on a crappy ASRock AM4 motherboard. Ended up replacing almost every part like PSU, RAM, and it kept happening, PC would just turn off in certain games or while downloading heavy games.
After there was nothing else to replace, I moved the setup to a decent B550 motherboard and that fixed it, haven't had any issues since.
buckeyebrownies13@reddit (OP)
Brutal, hopefully that’s not my issue
coolgaara@reddit
Finally, where I can actually help. I've had this exact same issue when I upgraded from 2700x to 5600x during COVID. I've spent 30+ hours literally trying every solution known to Google. Ended up returning it, because nothing worked. Put my old 2700x back? Boom, all good. A couple months later, it was available again and I so badly wanted it and bought it again like a fool. Guess what? Installed the new CPU and everything works fine. That's right, the CPU I had returned was defective.
So don't stress too much like I did because in the end, it could be as simple as a defective CPU. I'm sorry if this isn't the solution you're looking for but this was my experience.
buckeyebrownies13@reddit (OP)
Thanks dude. I keep leaning that way but the cpu passes the stress tests. Still not sure
ColdLaK@reddit
What is the event code in event viewer?
buckeyebrownies13@reddit (OP)
whea-logger event id 18
ColdLaK@reddit
I have the exact same CPU and have been getting this error message, along with crashes non stop since October 2nd. I would reach out to AMD for an RMA. I’m fairly certain it’s a defect with the 5800X3D based on my research
No_Package_6433@reddit
update amd chipset drivers too, i had the same problem with a 9700x
buckeyebrownies13@reddit (OP)
I updated the drivers. Issue was still there. Thank you though!
jhaluska@reddit
Crashes are super annoying to figure out.
I'd start with memtest86 and seeing if you have memory problems.
Then do a CPU stress test with Prime95.
buckeyebrownies13@reddit (OP)
I did a cpu test with cinebench, songle and multi core passed no issues
Scragglymonk@reddit
When you replaced the CPU, how did you clean the thermal paste off the cooler. How much and what pattern used for the new paste. Or did you skip that part ?
buckeyebrownies13@reddit (OP)
I used a thermal paste cleaner. I used a center dot with a small X.
SKtigercub88@reddit
If you have four ram sticks change to base settings if you have two set to xmp.
My computer kept crashing too and I couldn’t figure out why and then learned that if you have four sticks at xmp it will make it unstable and you’ll have to manually adjust timing. Since I don’t know how to do that I changed it to stock speeds and my crashing stopped.
DemSumBigAssRidges@reddit
In my experience, it has been this:
BSD and then restart = bad RAM.
No BSD but still restarts = bad (or not good enough) PSU.
BobbSacamano@reddit
Check your power plan settings. Make sure its on ultimate or high performance. Windows like to default on power saver which will lower your pcie, cpu, and ssd voltages causing crashes and restarts.
tokenwalrus@reddit
Can you check event viewer for shutdown events? Event IDs 41, 1074, 1076, 6005, 6006, 6008 are all related to shutdowns. Find the timestamp of when you experienced one of these shutdowns and see what the event says.
1counterparts@reddit
So this is just my personal experience
Years ago I bought a 5800X3D. The performance was SICK. I didn’t even over clock it. PC started black screening into a full reboot randomly. Event Viewer said something about CPU fatal error or something…installed old CPU, problem did not occur. RMAd 5800X3D, never had an issue with the replaced 5800X3D. Good luck!
Duedain@reddit
Where did you find a 5800x3d? I looked for months unsuccessfully until finding a 5700x3d which has been great, but wish I had found a 58.
buckeyebrownies13@reddit (OP)
A buddy in the industry
SirMaster@reddit
Does he have any more lol, I really want one...
Duedain@reddit
Nice! Congratulations. Should keep your AM4 system going for a long time to come. It REALLY boosted my gaming performance.
zam28@reddit
I installed an 5700x3d in an asrock b450 mobo, and have the same ptoblem, i reinstalled the cpu in another mobo gigabyte x470, and worked very well.
FunCelebration7751@reddit
Try upgrading your motherboard, there were problems with AsRock AM5 motherboards, even up to the B850 where they would just kill any of the X3D processors. Avoid AsRock and maybe get a X670E or an X870E.
What you’re experiencing is most likely 1. Power Delivery, or 2. Overheating. So add more Case Fans and switch to a bigger AIO if you can. It would also help to get a high quality PSU, see SPL’s PSU TierList for that.
Or if you cant do any of that, update your bios, i think the verified version that supports newer X3D chips are 3.12 and 3.15, depends on the motherboard
MouaTV@reddit
Clear CMOS. Look up how to do that for your specific motherboard. There may be two jumper pins you need to short. Otherwise, you can remove your CMOS battery, unplug PC from wall, then hold the power button for a few seconds (10 seconds should be fine).
Next, put your CMOS battery back in and boot into BIOS and load default settings. Boot up and see if that fixes it.
stinger32@reddit
I would turn it off, unplug it, and pull the CMOS battery to reset the BIOS. Turn on, check the BIOS settings, and save. One more restart and it should be good. I really do not know, but I like the idea. :-P
Full_Ad_3063@reddit
Ram o falta alimentacion?
simagus@reddit
Assuming you have the latest BIOS and nobody else is reporting issues with it (do check in case they are), you could have a hardware issue or a settings incompatibility.
The most likely thing is however the BIOS and/or it's current settings in relation to your new CPU.
You may have to tweak the voltage to get your 5800X3D running stable, but the fact it's crashing when you're not even under load is something best brought directly to the attention of ASRock customer support asap.
I would take this to their subreddit and make sure you are using the best BIOS revision for that CPU, as sometimes latest isn't the best for your specific system configuration.
If you have any error logs from or get any messages on your screen when the reboots happen, get those noted down and start looking for people with the same board and CPU combo.
buckeyebrownies13@reddit (OP)
Ok will do! Thank you for the reply. I will reach out to them now and cross post this into that sub