Is my Motehrboard dying or is too old for such certain things?

arajigar

Distinguished
Jun 1, 2016
213
3
18,715
Believe me, I am shocked. I have had this MOBO for years, and I thought it was rock solid...until now.

I am of those people that, once something fails, starts to not trusting on that something, specially with electronic components.

What is happening is that my GA-31-ESML rev 1.0 F10 Bios is not detecting my new GPU. I know, I know, this MOBO is almost 10 years old, but it does work perfectly until I put on it my new GPU, a GTX 1050.

It is not a problem of temps, or such things, my BIOS is up-to-date, and also drivers are. Temps in my system never raise over 25ºC, and the card is cold at touch. Is is perfectly sitted and the slot is clean as gold.

It all started with TWO BSODs, of unknown cause just because dumps were not written, that makes me think about the battery. Did a replacement of the battery and those BSODs disappear, but...

The problem is that the GPU now works 50% of the time. I mean: sometimes it is not regognised by the MOBO, the screen is set at 1024x760, and if you open device manager the error 43 appears, in the ohter side, when it works, resolution automatically appears at 1920x1080 and error 43 is gone.

This GPU has worked fine on other system with no problems, I did swaped it just in case and used in other PC with NO problems at all. Did several restarts, turned off, no problems. Same temps and almost the same GPU usage using the same games (the other pc is a core i5).

It seems that the PCI Ex slot is failing. Could it be possible?, also, I think that the problem is that the MOBO is not able to handle this GPU because of its age, or its chipset is too old, or the slot cannot supply enough energy... IDK, I´m not an expert, I'm just a computer shoddy.

Or should I buy a newer MOBO, as the ASROCK G41C.GS R2.0 is?.

Any help will be much apreciated. Thanks in anticipation.
 

arajigar

Distinguished
Jun 1, 2016
213
3
18,715
In fact the GPU works fine when ERORR 43 does not appear. I am now writing form this computer, and have played for a while with Doom 2016 and TitanFall with no problems at all. The sequence of the problem is as follows:

1.Turn OFF the system.
2.Turn ON the system. Start BOOT process. If all goes fine it takes only 15 secs. If it takes more than this time... is when I wnow that things are not working.
3.Screen is 1024x768, system works fine, except that resolution is not Ok and ERROR 43 appears in Device Manager.
4.I hit restart.
5.System turns on again, start booting, and it only takes 15 seconds, all is OK now, 1920x1080, no PROBLEM 43 present.
6.Turn it off. Wait some minutes, maybe an hour.
7.Return to point 2, point 3... sometimes it needs two reboots (the last time before I start writing this).

This GPU was only used for less than 30 days!!. And flawlessly... This MOBO probably is saying goodbye to me :(

As a last ditch effort, I'd go for a full uninstall of the drivers with DDU, then install the latest drivers from Nvidia's website.

If it's still misbehaving, then yeah, it looks like your motherboard might be suspect.

I forgot to say that that´s the first hing I did. Sorry for not saying it before. In fact I did Restart in Safe Mode-Run DDU -Run CCleaner- Reboot -Install latest drivers from Nvidia site after using their GPU DETECT online tool.
 

Rogue Leader

It's a trap!
Moderator


Ok thanks for the info. Can you perform a similar test on another system with your GPU in it?
 

arajigar

Distinguished
Jun 1, 2016
213
3
18,715
Did it. ;)

This GPU has worked fine on other system with no problems, I did swaped it just in case and used in other PC with NO problems at all. Did several restarts, turned off, no problems. Same temps and almost the same GPU usage using the same games (the other pc is a core i5).

It worked fine, tested with GPU and CPU hungry games and no problem. :( Maybe what I did write sounds like when you ask for the oppinion of another doctor when you have a bad diagnostic. In fact I am ashamed for doing so, but I was firing my last bullets trying to see if there was a solution. I think there´s no other than replacing my MOBO. :peur:
 

Rogue Leader

It's a trap!
Moderator


Ok got it, didn't realize you went through that whole sequence.

Before saying for sure its the board do you have access to another PSU to test? What PSU do you have right now? If its 10 years old like the motherboard theres a chance its not linking the extra power draw of the GPU.
 

Rogue Leader

It's a trap!
Moderator


Coolbox is a junk PSU brand. Could definitely be an issue.

Slots don't really fail, this sounds more like the board itself is having some sort of heat issue. You can try another slot in the board but I anticipate the outcome will be the same.
 

arajigar

Distinguished
Jun 1, 2016
213
3
18,715
Well, sadly, and unexpectedly, it´s the GPU which´s failing. I did more tests with the other PC and it started to do the same crazy things and the error 43 appeared too; I am writing now from the old PC with the GT740, and it works fine, IDK what happened, the GPU worked fine in the tests just to fail now, it´s only 1 month old, it was used in april and then the PC on which it was installed was unused until one week ago when I replaced this GTX1050 2 GB for its sister GTX1050 3GB. There were no surge problems,the PC was even disconnected from the wall, it worked fine; I repeat, before I wrote this post I did test it on the i5 PC with no problems at all, and today it just makes the same things: 800x600 (not 1024x768 resolution, sorry I was wrong in the other posts), error 43... appearing and disappearing randomly each time I turn off or restart the PC. Thanks for all your help!. I´m in shock.
 

arajigar

Distinguished
Jun 1, 2016
213
3
18,715
Sorry if I re-open this thread. I am not very confident about if my card will be RMA-ed. It´s still covered with warranty, but something is telling me that the problem will not be fixed, call me suspicious. So, I am asking again if... should the problem be caused by a bad GPUBIOS?, I mean, could it be corrupted or altered in a away it makes the card works intermittently? can a GPU BIOS be updated or fixed?. Thanks again in anticipation.
 


i wasn't aware that they even made a 2GB 1050. Where did you get the 2GB 1050 from? Where did you get the 3GB 1050? It is possible that you may have a counterfeit GPU(s), and this could be the cause of your issues.
 

King_V

Illustrious
Ambassador


The original 1050s were 2GB. Nvidia moving to 3GB for the 1050 is a relatively recent phenomenon.

If I remember correctly, the 3GB version's memory has less memory bandwidth, less L2 cache, and fewer ROPs, but compensated for by more SMs and CUDA cores, and slightly upped clock speeds on the GPU - which increases power draw and occasionally pushes the limit of the draw for the PCIe slot, for those that don't have a separate PCIe power connector.

See the comparison chart and overall review here:
https://www.tomshardware.com/reviews/geforce-gtx-1050-3gb-benchmarks,5654.html
 

Rogue Leader

It's a trap!
Moderator


Thats not the kind of thing that just happens. It takes specific software and effort to mess up the GPU BIOS. Thats not the issue.

 

arajigar

Distinguished
Jun 1, 2016
213
3
18,715
Sorry for the BUMP, but the problem was the PSU. Put another one with more horsepower and all is running fine for the last year. (GTX thus need minimum a 650 Watt Bronze to run on an old LGA 775 MOBO). Sorry for not having posted the solution before, I thought i wrote it the very moment I fixed all.
 
  • Like
Reactions: Rogue Leader

Rogue Leader

It's a trap!
Moderator
Sorry for the BUMP, but the problem was the PSU. Put another one with more horsepower and all is running fine for the last year. (GTX thus need minimum a 650 Watt Bronze to run on an old LGA 775 MOBO). Sorry for not having posted the solution before, I thought i wrote it the very moment I fixed all.


So, I was right the first time around. Coolbox PSU = junk.
 
  • Like
Reactions: arajigar