[SOLVED] Drivers from ASRock or from AMD?

Aug 29, 2020
26
5
35
0
My new build is getting more stable, but still restarts randomly. Yesterday it rebooted twice: once while my son was doing a project on Google Slides, and once overnight while nobody was near it (I woke up to the Steam login screen so I know it rebooted). I'm trying to exhaust all of the software/driver possibilities before I bring it to the local PC shop for hardware diagnosis. To that end, I've been trying to get the most up-to-date drivers. My mobo (ASRock B550m Pro4) and GPU (ASRock Phantom Gaming RX 5500 xt)seem to have different drivers on the Asrock website and AMD website. I'm looking at the chipset drivers on mobo and everything on the GPU.

Is there a general rule about where to get the drivers in this scenario?

AMD Ryzen 3600
ASRock B550m Pro4
ADATA XPG Spectrix S40g (512GB)
ASRock Phantom Gaming RX5500 XT
OLOy Warhawk DDR4 3600 (18-20-20-40)
EVGA 500BQ PSU (80+ Bronze)

https://pcpartpicker.com/b/cc4nTW
 
Last edited:

PC Tailor

Judicious
Ambassador
Doesn't really matter. If it's for the card, just get them from AMD as you'll be certain that they'll pretty much always be up to date.

I've just looked at your previous post, I'd be surprised for drivers to cause the manner you describe. Yes I would always update drivers etc. as first port of call, but it's not always like driver issues to cause sudden restarts in that way, they are more often likely to cause stop errors or hangs for example. Not to say it never happens, it does, just might not be the only thing.

Once you have ensured all drivers are up to date, I'd verify temperatures under load, not at low loads first just to make sure. Following that I would also just check what BIOS version you are running. Sometimes BIOS patches can help with hardware, just if the issue is a hardware fault itself, you wouldn't want the PC restarting during a BIOS update.

Cheap PSU issues seem to be the main suggestion. I didn't go too cheap there, and 500W SEEMS adequate.
Correct, but not cheap doesn't always mean good. The BQ is not a great quality PSU. Certainly fine, but not great. And could well be the issue. The BQ is a mediocre PSU and is a bit of a mixed bag with how well they might run in a system. But ideally you'd keep them on lower end systems (maybe mid tier at a push).

Quite often once software has been eliminated, random restart culprits from general experience tend to go in this order:
PSU > GPU > RAM > MB > CPU.
 

PC Tailor

Judicious
Ambassador
Doesn't really matter. If it's for the card, just get them from AMD as you'll be certain that they'll pretty much always be up to date.

I've just looked at your previous post, I'd be surprised for drivers to cause the manner you describe. Yes I would always update drivers etc. as first port of call, but it's not always like driver issues to cause sudden restarts in that way, they are more often likely to cause stop errors or hangs for example. Not to say it never happens, it does, just might not be the only thing.

Once you have ensured all drivers are up to date, I'd verify temperatures under load, not at low loads first just to make sure. Following that I would also just check what BIOS version you are running. Sometimes BIOS patches can help with hardware, just if the issue is a hardware fault itself, you wouldn't want the PC restarting during a BIOS update.

Cheap PSU issues seem to be the main suggestion. I didn't go too cheap there, and 500W SEEMS adequate.
Correct, but not cheap doesn't always mean good. The BQ is not a great quality PSU. Certainly fine, but not great. And could well be the issue. The BQ is a mediocre PSU and is a bit of a mixed bag with how well they might run in a system. But ideally you'd keep them on lower end systems (maybe mid tier at a push).

Quite often once software has been eliminated, random restart culprits from general experience tend to go in this order:
PSU > GPU > RAM > MB > CPU.
 

Math Geek

Champion
Ambassador
i tend to prefer manufacturer releases first. they should be designed to go along with any specifics for that card/mobo.

but if they are causing problems, i have no problem going to amd to get others to try. they should both work and the newest is not always the best.

sometimes the new ones are what can cause issues and its worth going back to an older stable version.
 
Reactions: PC Tailor
Aug 29, 2020
26
5
35
0
Doesn't really matter. If it's for the card, just get them from AMD as you'll be certain that they'll pretty much always be up to date.

I've just looked at your previous post, I'd be surprised for drivers to cause the manner you describe. Yes I would always update drivers etc. as first port of call, but it's not always like driver issues to cause sudden restarts in that way, they are more often likely to cause stop errors or hangs for example. Not to say it never happens, it does, just might not be the only thing.

Once you have ensured all drivers are up to date, I'd verify temperatures under load, not at low loads first just to make sure. Following that I would also just check what BIOS version you are running. Sometimes BIOS patches can help with hardware, just if the issue is a hardware fault itself, you wouldn't want the PC restarting during a BIOS update.


Correct, but not cheap doesn't always mean good. The BQ is not a great quality PSU. Certainly fine, but not great. And could well be the issue. The BQ is a mediocre PSU and is a bit of a mixed bag with how well they might run in a system. But ideally you'd keep them on lower end systems (maybe mid tier at a push).

Quite often once software has been eliminated, random restart culprits from general experience tend to go in this order:
PSU > GPU > RAM > MB > CPU.
Hey, thanks so much for such a thorough reply.

I ran stress tests on the CPU (CPU-Z) and GPU (Furmark) without a failure, plus ran a memory scan (DRAM calculator for Ryzen->memtest) with no errors reported. My temps always look OK per RyzenMaster and ASRock Tweak. Is there an easy way to check the PSU at home before I have to pay the local shop?
 
Aug 29, 2020
26
5
35
0
i tend to prefer manufacturer releases first. they should be designed to go along with any specifics for that card/mobo.

but if they are causing problems, i have no problem going to amd to get others to try. they should both work and the newest is not always the best.

sometimes the new ones are what can cause issues and its worth going back to an older stable version.
Thanks for this. This all makes sense. Maybe I'll try the ASRock version for a day and see how things go. If I get more restarts I can try AMD. If that doesn't work, off to the shop I go.
 

PC Tailor

Judicious
Ambassador
Hey, thanks so much for such a thorough reply.

I ran stress tests on the CPU (CPU-Z) and GPU (Furmark) without a failure, plus ran a memory scan (DRAM calculator for Ryzen->memtest) with no errors reported. My temps always look OK per RyzenMaster and ASRock Tweak. Is there an easy way to check the PSU at home before I have to pay the local shop?
Not effectively. One can monitor voltages on each rail through software such as HWInfo but these figures are not always accurate or precise, but they can be a good indicator of anything obviously wrong.

Second to that one can always test the voltage on rails manually by disconnecting the PSU, powering it up and then testing each rail, assuming you can work safely with electronics. Problem with this method is it doesn't actually stress the system in any way, and most PSU issues occur under load, not idle. So you don't see the full picture.

The best and most effective way of testing a PSU is the simplest, replace it with another known working unit and see if the issue persists. This is what a shop should do as well.

Try the software first, and check your BIOS version is up to date, then lets see :)
 
Aug 29, 2020
26
5
35
0
Try the software first, and check your BIOS version is up to date, then lets see :)
Got it, and thanks again. Yes, I updated the BIOS (even though I hate updating the BIOS) to the latest available version. I wish I could figure out what use scenario was triggering the failures. At this point I am making little changes, then waiting a day or so with fingers crossed, but I can't say for sure "It crashes on Minecraft" or with a particular game or program or stress test. I thought it was figured out until the two crashes yesterday. So even if I put in a new PSU I'll just have to wait around and see if that was the fix I guess.

I also have an old HDD in there from an old system for bulk storage. I have a new SSD coming today to replace it. Once I make that swap I might just try fully resetting Windows (10, Home, 64 bit) and starting fresh.

Thanks for your suggestions. Here I thought this build would just be a little weekend project! LOL
 
Last edited:

DSzymborski

Champion
Moderator
Got it, and thanks again. Yes, I updated the BIOS (even though I hate updating the BIOS) to the latest available version. I wish I could figure out what use scenario was triggering the failures. At this point I am making little changes, then waiting a day or so with fingers crossed, but I can't say for sure "It crashes on Minecraft" or with a particular game or program or stress test. I thought it was figured out until the two crashes yesterday. So even if I put in a new PSU I'll just have to wait around and see if that was the fix I guess.

I also have an old HDD in there from an old system for bulk storage. I have a new SSD coming today to replace it. Once I make that swap I might just try fully resetting Windows (10, Home, 64 bit) and starting fresh.

Thanks for your suggestions. Here I thought this build would just be a little weekend project! LOL
Wait, did you carry over Windows 10 install from your previous build? Windows 10 isn't intended to be modular in this way unless you have a very specific Windows-to-go install (which you would have to quite intentionally do). Best practice is a full wipe-and-reinstall on platform upgrades.

I certainly would have done it before worry about the BIOS. I agree with you about trying to avoid that one.
 
Aug 29, 2020
26
5
35
0
Wait, did you carry over Windows 10 install from your previous build? Windows 10 isn't intended to be modular in this way unless you have a very specific Windows-to-go install (which you would have to quite intentionally do). Best practice is a full wipe-and-reinstall on platform upgrades.

I certainly would have done it before worry about the BIOS. I agree with you about trying to avoid that one.
No, I did a clean install of Windows to start. :) I'm just wondering if something might have gotten corrupted in all of my tinkering with drivers and whatnot. None of the driver installer apps gave me the option to erase the previous driver and do a "clean" install of the new one, so I am wondering if there are some broken bits of driver or registry entries floating around. I even tried the AMD Cleanup Utility, but it seems that deleted my current GPU driver (when I checked after running the utility, it had reverted to a Microsoft driver from 2006).
 
Aug 29, 2020
26
5
35
0
UPDATE: I went with the AMD.com drivers, as the ASRock page even suggested doing so.

I also went nuclear and upgraded my PSU from the EVGA BQ500 to a Super Flower Leadex III 650. Overkill for the current system, but with a 10 year warranty it should last me through plenty of future upgrades. Played a few games, left it on overnight, played a few more games today and so far no random restarts (knock, knock). I found out I was outside the return window for the EVGA power supply, so this was an expensive learning experience, but at least it seems I have some stability now.
 

ASK THE COMMUNITY

TRENDING THREADS