Old Asus STRIX z490-E gaming rig has become unstable - how to determine if it's motherboard, CPU or something else?

Jack Of Owls

Jack Of Owls

Weaksauce
Joined
Jan 3, 2021
Messages
111
Starting a couple months ago, my system has become unstable. Large downloads get easily corrupted, and sometimes unraring gives CRC errors, no matter which SSD drives. I used to get Windows 10 BSODs but since updating the OS that seems to have stopped but the other issues remain. Using both Windows memory diagnostic and MemTest86 causes many RAM errors immediately, even with switching out the two 16 GB sticks, testing different slots or testing each individually which makes me think it's not a ram issue (what chances are there that both sticks would go bad at once?). I suspect it's a bad motherboard/memory controller or cpu but it seems a daunting task to find out exactly which. Feel like I'd be chasing my tail.

I have no spare z490 boards or 10th gen Intel chips to test with. Is it time to finally upgrade to later generation motherboard/cpus or can this rig be saved by just replacing the motherboard or cpu (I've had it about 5 years and I really like it)? The trick is finding out which. Is there an easy way? Ideally, I would love to just replace the board or chip if I knew exactly which is the problem.

My specs:
CPU: Intel i9-10850K 10-core 3.6 GHz (5.2 GHz turbo) 20MB Cache
MB: Asus ROG STRIX z490-E Gaming Motherboard
RAM: G.Skill Ripjaws VF4-4000C18D-32GVK DDR4-4000MHz CL18-22-22-42 1.40V 32GB (2x16GB)
GPU: EVGA GeForce RTX 3070 XC3 BLACK GAMING, 8GB GDDR6, iCX3 Cooling, ARGB LED, LHR, 08G-P5-3751-KL
PSU: Super Flower Leadex III 750W 80+ Gold, Three-Way ECO Mode Fanless, Silent & Cooling Mode, FDB Fan, SF-750F14HG
OS: Microsoft Windows 10 Professional 64-bit (Version 21H2, Build 19044.1766)
EVGA 280mm liquid cooler
Various SSD drives, some SATA some M.2
 
First things I'd do is set the mem to a much lower speed and update the bios. I'd also switch to a hardline if using wifi or vice versa to eliminate the internet adapter as the cause.
 
Yes, start by running JEDEC speeds on the RAM.

The general answer to your question in the title is: you can't. If CPU or motherboard is broken you can't tell which it is. Even RAM - not every memtest error is defective RAM, and not every defective RAM error is detected by memtest (try superpi).
 
Furious_Styles said:
First things I'd do is set the mem to a much lower speed and update the bios. I'd also switch to a hardline if using wifi or vice versa to eliminate the internet adapter as the cause.
Click to expand...
Just had a BSOD an hour or two ago so I think my PC is nearing its end of life. I used to have XMP OC profiles set in the z490-E's BIOS (my Gskill sticks supposedly supported it) but started having issues about a year and a half ago so I set my memory to default values around that time. Updated the MB's BIOS about 2 weeks ago to latest version. Yes, I've generally only used a hardline for internet since I got the motherboard 4-5 years ago since I get twice the throughput compared to wi-fi. I'm thinking my old rig is just wearing out so I'm heavily leaning towards just getting a motherboard bundle for around $400 if I can find a good sale. I don't need the latest PC tech; I'm usually a generation or two behind and I'm okay with that. Thinking about going back to AMD since all that heat generated by my old 10th gen Intel cpu is an annoyance, even with liquid cooling. I do game a lot so that's a consideration. Thx for the reply.
 
uOpt said:
Yes, start by running JEDEC speeds on the RAM.

The general answer to your question in the title is: you can't. If CPU or motherboard is broken you can't tell which it is. Even RAM - not every memtest error is defective RAM, and not every defective RAM error is detected by memtest (try superpi).
Click to expand...
Yeah, I figured there's no easy way to troubleshoot the motherboard from the cpu. Looks like it's time to chuck this old motherboard/cpu and get something newer.
 
Jack Of Owls said:
Just had a BSOD an hour or two ago so I think my PC is nearing its end of life. I used to have XMP OC profiles set in the z490-E's BIOS (my Gskill sticks supposedly supported it) but started having issues about a year and a half ago so I set my memory to default values around that time. Updated the MB's BIOS about 2 weeks ago to latest version. Yes, I've generally only used a hardline for internet since I got the motherboard 4-5 years ago since I get twice the throughput compared to wi-fi. I'm thinking my old rig is just wearing out so I'm heavily leaning towards just getting a motherboard bundle for around $400 if I can find a good sale. I don't need the latest PC tech; I'm usually a generation or two behind and I'm okay with that. Thinking about going back to AMD since all that heat generated by my old 10th gen Intel cpu is an annoyance, even with liquid cooling. I do game a lot so that's a consideration. Thx for the reply.
Click to expand...
Yeah you need a set of spares to figure out whether it's the cpu/mobo/mem typically. For the hell of it you can run the intel diagnostic test, but I bet it passes.
https://www.intel.com/content/www/us/en/download/15951/intel-processor-diagnostic-tool.html
 
Furious_Styles said:
Yeah you need a set of spares to figure out whether it's the cpu/mobo/mem typically. For the hell of it you can run the intel diagnostic test, but I bet it passes.
https://www.intel.com/content/www/us/en/download/15951/intel-processor-diagnostic-tool.html
Click to expand...
Got another BSOD today and also just had a weird issue with my Firefox browser not able to load anymore, complaining of not being able to send an error report (prior to this over the last month I got many crashing tabs, further evidence of instability due to something failing). Had to reboot to get Firefox back. I'll check that benchmark out just for shits and giggles but I find software stability tests very unreliable over the years (i.e. I used OCCT yesterday and all tests passed so... pretty useless). I'm planning my budget build now and I'm considering a Ryzen 5700X3D with an Asrock X570S PG Riptide board. I think I might be able to keep my Gskill F4-4000C18D-32GVK sticks (but maybe not because the the board is not on gskill's QVL list but the X570 chipset is supported in other brands) and EVGA water cooler.

Like I said, I tend to go the budget route and will be a couple generations behind what's current but at least I move up from PCIe 3.0 to PCIe 4.0. Asrock X570S PG Riptide has all the features I need. I've had enough of the once great and reliable Asus. That's two straight Asus boards that failed on me in the last few years.
 
Last edited:
Jack Of Owls said:
Got another BSOD today and also just had a weird issue with my Firefox browser not able to load anymore, complaining of not being able to send an error report (prior to this over the last month I got many crashing tabs, further evidence of instability due to something failing). Had to reboot to get Firefox back. I'll check that benchmark out just for shits and giggles but I find software stability tests very unreliable over the years (i.e. I used OCCT yesterday and all tests passed so... pretty useless). I'm planning my budget build now and I'm considering a Ryzen 5700X3D with an Asrock X570S PG Riptide board. I think I might be able to keep my Gskill F4-4000C18D-32GVK sticks (but maybe not because the the board is not on gskill's QVL list but the X570 chipset is supported in other brands) and EVGA water cooler.

Like I said, I tend to go the budget route and will be a couple generations behind what's current but at least I move up from PCIe 3.0 to PCIe 4.0. Asrock X570S PG Riptide has all the features I need. I've had enough of the once great and reliable Asus. That's two straight Asus boards that failed on me in the last few years.
Click to expand...
Are you positive it's not one of your data drives? It could be the mobo but getting corrupted downloads could also be a bad sata drive or cable. Always gotta check the easy stuff just in case. Also what error do you get on the BSODs?
 
Furious_Styles said:
Are you positive it's not one of your data drives? It could be the mobo but getting corrupted downloads could also be a bad sata drive or cable. Always gotta check the easy stuff just in case. Also what error do you get on the BSODs?
Click to expand...
I tend not think it's a data drive because I tested multiple SSDs for my downloads, including a brand new HP 4TB NVNe. Today I discovered I can't even copy from one drive to another without getting corrupted files on the destination drive. Also, when I run the memory diagnostics, I'm bypassing all my data drives (except for 1) and booting into the diagnostics before windows. I get the standard Windows 10 blue screen hardware failure screen with the frowny face followed by an instant reboot. No, I'm fairly certain my motherboard/cpu is falling because it's getting worse by the day. At least I have a new project to look forward to over the next week with the upgrade. I tend to not upgrade to a new motherboard/cpu unless the old one failed or is failing, unless I suddenly get an unexpected financial windfall ;)

Edited: one thing I didn't consider is that my boot SSD is failing but that wouldn't explain why I instantly get all those memory errors when I boot fresh into MemTest86 without even touching my boot drive.
 
Last edited:
pendragon1 said:
i said bsod code, not error logs....
Click to expand...
pendragon1 said:
i said bsod code, not error logs....
Click to expand...
I've never done that before and didn't even know that was a thing you could do. But I just disabled auto-restart after a system failure so I'm assuming a 10-digit code comes up after a BSOD and that might help diagnose the issue. I'll give it a try next time I get BSODed.
 
This is the code needed.

9b313b0b-ea7c-400a-92cb-f725d84c83ba.png
 
It's nice to know that you got a segfault in the kernel, but without a backtrace that is not useful.
 
Jack Of Owls said:
Using both Windows memory diagnostic and MemTest86 causes many RAM errors immediately, even with switching out the two 16 GB sticks, testing different slots or testing each individually which makes me think it's not a ram issue (what chances are there that both sticks would go bad at once?).
Click to expand...

Got any ram from another system you can try? (Switching sticks wasn't clear if you just swapped from the same machine)

If you can't get a clean ram test by running JEDEC (slow) timings and setting an appropriate memory voltage, especially if you put in fresh ram, it's likely your board or your cpu; or maybe your power supply. Not much help.

Two ram sticks that were made at the same time and subject to the same use conditions failing at the same time is actually pretty reasonable.
 
toast0 said:
Got any ram from another system you can try? (Switching sticks wasn't clear if you just swapped from the same machine)
Click to expand...
No, I don't. I wanted to transfer this ram over to the replacement board but I'm concerned now it's either incompatible with the motherboard I'm leaning towards getting or it's failing. I saw a newegg deal today for the Ryzen cpu I've been eyeing that comes with free ram (Gskill 32GB 3600 MHz) so I might go for it just to be safe
 
Jack Of Owls said:
No, I don't. I wanted to transfer this ram over to the replacement board but I'm concerned now it's either incompatible with the motherboard I'm leaning towards getting or it's failing. I saw a newegg deal today for the Ryzen cpu I've been eyeing that comes with free ram (Gskill 32GB 3600 MHz) so I might go for it just to be safe
Click to expand...

I'd probably get a whole bundle given the deals. Sounds like am4? You can try testing the old ram in the new system and the new ram in the old system and see if you can make a guess at which is which and decide if you want to make a second box maybe?

Flakey systems suck; dealing with the kiddo's desktop being super crashy for a while... but mine is worse, no memory errors in tests, but crash codes are most likely bad ram. :(
 
toast0 said:
I'd probably get a whole bundle given the deals. Sounds like am4? You can try testing the old ram in the new system and the new ram in the old system and see if you can make a guess at which is which and decide if you want to make a second box maybe?

Flakey systems suck; dealing with the kiddo's desktop being super crashy for a while... but mine is worse, no memory errors in tests, but crash codes are most likely bad ram. :(
Click to expand...
Yeah, I was looking at a Ryzen 5700X3D cpu with an Asrock X570S PG Riptide AMD X570 AM4 board as my new possible setup. The Ryzen comes with the free ram. That would be great if my current instability problems turned out to be bad ram then I could just buy new sticks and have a second PC (actually a fourth; I've got like 3 computers in storage lol) that's almost as powerful as the new one I intend to put together. I'm currently checking the slickdeals forums just in case a better bundle comes along.
 
As an Amazon Associate, HardForum may earn from qualifying purchases.
Jack Of Owls said:
Edited: one thing I didn't consider is that my boot SSD is failing but that wouldn't explain why I instantly get all those memory errors when I boot fresh into MemTest86 without even touching my boot drive.
Click to expand...
Have you tried testing one DIMM at a time with memtest86? Usually bad ram is just one stick, and sometimes a bad board is one ram socket and bad CPU can be one memory channel.
 
zandor said:
Have you tried testing one DIMM at a time with memtest86? Usually bad ram is just one stick, and sometimes a bad board is one ram socket and bad CPU can be one memory channel.
Click to expand...
Yeah, that was one of the first troubleshooting steps I took. Result: immediate (and many) errors, with both or either sticks/slots. I'll probably do one more memtest86 test before I order the new replacement components though.
 
Today I got a BSOD and the error was: Stopcode: memory management

So... maybe a bad memory controller on the motherboard?
 
CAD4466HK said:
I got the same error and had 2 bad sticks of RAM.

https://hardforum.com/threads/how-can-brand-new-memory-go-bad-within-weeks.2026268/#post-1045948772

Every kit of RAM I've ever owned had a lifetime warranty, RMA it, a cheap kit of DDR 4 only costs $36 if you can't have down time.
Click to expand...
My computer is basically hopping a marathon on one leg but I think i might be able to keep kicking it in the @ss to keep it moving to the finish line until a week or two when I get the new components ;) well, at least I learned something new - 2 different ram sticks can go bad within a week/day, though I still suspect it's relatively rare. I'm wondering if I can RMA it though I'd have to make certain they're dead or dying. When I get the new components, I'll test them in the new motherboard.

I have a question though for the gurus here: if the sticks are still good, is it possible to use them in the new Asrock motherboard I'm planning on buying WITH the new ram modules if they're the same brand (G.skill Ripjaws) and seemingly almost the same model, except my current sticks are rated for 4000 MHz (VF4-4000C18D-32GVK DDR4-4000MHz), and the new G.Skill sticks are rated for 3600 MHz (F4-3600C18D-32GVK DDR4-3600 MHz)? I was thinking what a nice bonus it would be if I can just add them to get 64GB of system ram out of all this at the end of my PC tech trials and tribulations here :D
 
Furious_Styles said:
Might be worth it to buy some cheap memory and see if it fixes it. DDR4 is inexpensive now.
Click to expand...
Yeah, just bought a couple of G.Skill sticks (32GB total) for around $50 (I paid over $100 4 years ago for nearly the same sticks, just 4000MHz instead of 3600MHz!) as my first purchase in the upgrade. Will suck it & see if it runs stable in my old board, and if it doesn't I know the problem is the motherboard/cpu then I'll buy the rest of the components. My first experiences with incompatible RAM in the 1990s usually had symptoms of constant program CTDs, if it booted at all. I get something similar with Firefox tabs here plus the BSOD errors. That would be so great if I could get a few more years out of this setup with maybe just upgrading the GPU in a year or two.
 
Moparzrule said:
How old is the power supply? Unstable voltages can cause issues.
Click to expand...
Practically brand new; less than a year - a Super Flower Leadex III 750W 80+ Gold. However, I did have strange issues one day where the fan started noisily chugging along, and literally a day or two before I was authorized to RMA it back to the manufacturer (I sent them a video; they said it sounded like a failing ball bearing), it corrected itself. How bizarre.
 
Jack Of Owls said:
Practically brand new; less than a year - a Super Flower Leadex III 750W 80+ Gold. However, I did have strange issues one day where the fan started noisily chugging along, and literally a day or two before I was authorized to RMA it back to the manufacturer (I sent them a video; they said it sounded like a failing ball bearing), it corrected itself. How bizarre.
Click to expand...
Gotcha. I had never heard of that PSU brand but they seem decent. I mean, a fan bearing failure after a year kinda make me think a little low quality though. I didn't read the rest of the posts, but if you tried replacing the ram already I would honestly thing PSU would be the next most likely culprit. If you didn't try replacing the ram, did you try increasing the voltage to 1.45v at all? Sometimes after some age, they just need a little more juice.
 
Moparzrule said:
Gotcha. I had never heard of that PSU brand but they seem decent. I mean, a fan bearing failure after a year kinda make me think a little low quality though. I didn't read the rest of the posts, but if you tried replacing the ram already I would honestly thing PSU would be the next most likely culprit. If you didn't try replacing the ram, did you try increasing the voltage to 1.45v at all? Sometimes after some age, they just need a little more juice.
Click to expand...
Yes, I realized they lose a little juice when about a year or two ago I could no longer use the XMP 4000MHz profile in my BIOS without massive instability so I had to bump it down to 3200 MHz to get it stable again. I'll try the bumping up of voltage if the new ram sticks I ordered don't give joy. Regarding cheap PSUs, I remember taking advice from an OEM builder of systems and purchasing the inexpensive PSU he recommended that had excellent voltages on those 12v volt rails, and having it die within a year. Then I bought another one of the same model, and THAT died within a year. Only difference was it left a scorch mark on my Abit motherboard this time. I figured if I bought another one of that brand, my poor motherboard would have been destroyed. A frequent tip-off of these cheapjack PSUs is you can hear some dislodged component rattling around inside when you shake it ;)
 
pendragon1 said:
theyve been a go to brand for years and are an OEM for others too.
Click to expand...
Personally, I would have come up with a better name for a company or a brand of power supplies than "Super Flower" but they do seem to be highly regarded.
 
Last edited:
Jack Of Owls said:
Yes, I realized they lose a little juice when about a year or two ago I could no longer use the XMP 4000MHz profile in my BIOS without massive instability so I had to bump it down to 3200 MHz to get it stable again. I'll try the bumping up of voltage if the new ram sticks I ordered don't give joy. Regarding cheap PSUs, I remember taking advice from an OEM builder of systems and purchasing the inexpensive PSU he recommended that had excellent voltages on those 12v volt rails, and having it die within a year. Then I bought another one of the same model, and THAT died within a year. Only difference was it left a scorch mark on my Abit motherboard this time. I figured if I bought another one of that brand, my poor motherboard would have been destroyed. A frequent tip-off of these cheapjack PSUs is you can hear some dislodged component rattling around inside when you shake it ;)
Click to expand...
Honestly you are lucky, everytime I've ever had a PSU go it took out the motherboard.
 
You must log in or register to reply here.
Back
Top