Faulty Gpu/Mobo ?

Toggle sidebar Toggle sidebar
A

alfac95

Reputable
Jan 3, 2015
13
0
4,510
My graphics card, an MSi R9 270 has died for the 3rd time after respective RMA processes, and it dies in the same fashion every time. I've kept a watch on the temps which were fine, and they don't really matter as the card died this once after 2 mins of booting and getting in-game. I get the Grey Screen of Death at first, and the next time I boot and open a game, my card artifacts out and the display turns black. The fans are spinning, tried switching PSUs, got 2 repairs and one replacement, but to no avail. I'm not sure what's killing the card, a faulty PCIe slot, the PSU (Unlikely, had a coolmaster 600w before, now a new 550 corsair) or something entirely else. The system runs perfect on the integrated display, manages 10-20 fps on gta V for prolonged hours (test purposes) whithout any mishaps; which makes me believe it's somethinf about the PCIe slot or the GPU. What exactly is the culprit here?

System:- i5 3330 @3.2, 8 gigs 1333 mhz ram, MSi R9 270, 1tb hdd@ 7200, 550w corsair vs psu (not cool but was on a low budget, and thought it's fair for my rig), Gigabyte GA-H61M-s1 Mobo - All at stock clocks
 
Solution
brandonclone1
It's good troubleshooting to replace parts as needed. With you GPU and PSU replaced, I agree with you that your PCIe slot or the motherboard itself is your next suspect. Basic mobo troubleshooting is to check for blown capacitors, use a multimeter if you have one (http://www.ehow.com/how_8723689_test-motherboard-multimeter.html). And as always, run Windows updates and AMD driver updates.

If I were you, I'd shell out for one of the new RX cards since they are affordable and pretty powerful. If the same problems occurs with a completely different card model, you've narrowed it down to a faulty mobo and you can trash Gigabyte for selling you a POS motherboard :)
Sort by date Sort by votes
brandonclone1

brandonclone1

Honorable
Mar 26, 2014
94
2
10,665
It's good troubleshooting to replace parts as needed. With you GPU and PSU replaced, I agree with you that your PCIe slot or the motherboard itself is your next suspect. Basic mobo troubleshooting is to check for blown capacitors, use a multimeter if you have one (http://www.ehow.com/how_8723689_test-motherboard-multimeter.html). And as always, run Windows updates and AMD driver updates.

If I were you, I'd shell out for one of the new RX cards since they are affordable and pretty powerful. If the same problems occurs with a completely different card model, you've narrowed it down to a faulty mobo and you can trash Gigabyte for selling you a POS motherboard :)
 
Upvote 0 Downvote
Solution
geofelt

geofelt

Titan
Oct 9, 2006
54,295
4,145
174,240
neither psu is considered as high quality.
http://www.tomshardware.com/forum/id-2547993/psu-tier-list.html
That said, it is not likely to be the cause of your problem.

When you get a new card from the rma process, it seemed to work for a while.
Then it fails.
That suggests to me that something is causing damage to your cards.
Possibly the motherboard is at fault and is delivering higher current to the pcie slot than it can handle.
See if there is a bios update to the motherboard that addresses the issue.
DO NOT flash the bios on just speculation.
A failed flash can brick the motherboard.

Or, the design of the card draws more than the 75w spec. RX480 had this issue.
Unlikely too.

If these are new parts, still under warranty, I would try to return both the motherboard and the graphics card and try different brands.










 
Upvote 0 Downvote
A

alfac95

Reputable
Jan 3, 2015
13
0
4,510
Both the PSUs aren't any true gold performers for sure, but I considered my rigs total TDP consumption and settled down for them as a mismatched overkill psu would be a waste for mere testing purposes. Clean installed drivers without any overlaps to avoid complications, but it is hardware related by now I've known. The graphics card still is in a year's warranty span after 2 years, glad that MSi doesn't ask one single question and does it's best everytime, but I'm afraid the rest of those components are long out of warranty services, bought them probably 5+ years ago :p So before I go for a current gen hefty upgrade, I'm trying to patchwork this rig if it has got any chance, because it still served me strong enough. Thanks for the replies, I'll now send the card for RMA and consider arranging an alternate mobo, this looks fine (The capacitors are intact, nothing leaks or swells), probably the slot gave up
 
Upvote 0 Downvote
geofelt

geofelt

Titan
Oct 9, 2006
54,295
4,145
174,240
When you get the new card, can you test it first in another pc?
That way, you will know for certain that it is not the card and narrowed it down to the motherboard or psu if it fails again.
And... put a mark on the card you return so you know it is truly a replacement.
 
Upvote 0 Downvote
A

alfac95

Reputable
Jan 3, 2015
13
0
4,510
Yes I could do that for sure, although most likely I've zero downed to it being the mobo due to factors like having tested on 2 supplies, one being less than 2 months old, resulting in similar deaths, all temps being genuine etc. The total power consumption should not exceed 350 watts at any circumstance at all under it's extreme peak, which is sufficiently being met. The integrated chip gives endless hours of stable operations (Considering that while on intel hd the power consumption drops). Everything starts and ends after the card sits inside. The mobo is already, like 5-6 years old at it's least, the psu is new, the gpu is "new", processor is functioning fine yada yada there's no harm in testing for sure tho. I really am tired of doing the RMA process tbh again and again, which is hell time consuming and seems pointless if the piece is going to collapse every once within a week; secretly hoping it's my PCI slot which went rogue
 
Upvote 0 Downvote
You must log in or register to reply here.

TRENDING THREADS

Latest posts

Moderators online

Share this page

COMPANY
RESOURCES
FOLLOW
Top Bottom