Is my 680M dying?

Hey foks

Having trouble with my 680M. It is refusing to run well under load.

If I play a game or the gpu-z render test for more than a minute the system simply freezes and I have to force it off. The gpu is at stock clocks and voltage.

Another thing I noticed from gpu-z is the 680M is only running at pci-e 2.0 x8 not x16. It will cycle 1.1-2.0 depending on load as normal but will not go to x16 like is always used to.

The laptop has no problem driving the 920xm even in Prime 95 at full load (3.7ghz roughly 180W). Does this mean my 680M is dying?

After one incident I had the ominous blinking scroll and caps and solid num which apparently indicates faulty gpu primarily or secondarily motherboard according to Alienware documentation (M15x) yet after a simple power drain I was up and away again booting into windows 7 as normal.

Any insight would be great! Also note I haven't had any artifacts or lines on the screen simply the card will not run at 3d clocks without quickly freezing the system (not even BSOD) I am thinking the vrms on the card are toast yet not sure.

Anyways this has crazy timing as there is a 980M on it's way to me right now. If the 680M is dead I have a replacement BUT if the motherboard is the culprit I would rather not place a new card in and throw away a perfectly good 680M that can be sold and put to good use elsewhere as well as risk breaking the new card when the mobo finally decides to completely fail.

I am leaning towards faulty gpu and next week the 980M will reveal more but just wanted to get other opinions

Thanks!

 

It sounds like the motherboard to me.

 

Hi Triturbo! Thanks for your thoughts You could well be right.

Strange though. I rarely game tbh. When I do it is intensive but other than for the occasional bench I always run on as little voltage as possible.

The cpu however I ran for months overvolted yet the power delivery for it is fine. Shouldn't those vram be more likely damaged than the ones feeding the gpu or are the vrm's shared.

Ironically annoying as that might be a new motherboard is cheaper than a dead 680M. In fact roughly half the cost lol

 

Having a good think about this. I remember spilled ALOT of orange juice onto the laptop a year ago. I was quick to act but the juice did get into the chassis.

I replaced the keyboard with a new one, cleaned up the chassis beneath the keyboard as best I could and forgot about it.

Could this have slowly damaged the motherboard?

 

Usually liquid damage that causes issues is evident if you look closely at the motherboard. If there are spots of corrosion on the board or near the GPU area then that is probably your issue. Corroded traces that are falling apart can for sure cause lots of issues which most of the time are inconsistent (just like in your case). I think it is worth to disassemble and make a very good inspection of the board. In these situations, throwing parts at it is not cost effective so you have to find what the cause before repairing.
Also note that corrosion from liquid damage doesn't happen overnight, in fact it is quite a long process and depends on the liquid that was spilled.

 

What he said. If you haven't cleaned your MoBo thoroughly with rubbing alcohol, the chances for bad outcome increase dramatically. A lot of goo can be "locked" underneath the components, paving their way to death, especially when we talk about sugarish drinks.

 

Well I have decided to get a new motherboard. It isnt expensive. No more than the ram in fact lol.

I dont see the point installing the valuable 980M into a faulty motherboard and take unnecessary risks.

Besides a new motherboard will see the machine out with no more issues (unless I get clumsy again)

Sent from my SM-A500FU using Tapatalk

 

Well it is pretty much dead now (graphically) started to stay permanently at 3d clocks. I took it down to 0.98V and then the bus speed switched to pcie 1.1 x8 and now once into windows screen goes black.

Only on safe mode where 2d is forced can i get picture.

That died quickly!

Sent from my SM-A500FU using Tapatalk

 

Do you have another laptop to try it in?

To disqualify that it's a mobo issue.

 

I am hoping 980M arrives soon. I can test that and see how it behaves. No spare mxm laptop though.

The motherboard is about 6 years old though. A refresh isnt a bad idea even if the gpu is the culprit. Considering 980M cost 5 times as much

Sent from my SM-A500FU using Tapatalk

 

Well the 980M arrived. The card although not quite as bad as the 680M is still throttling heavily on the core in games and often wont go over pcie 1.1.

If I try and use dual adaptors performance gets even more erratic with changing brightness level flashing on the screen.

Nope the mobo is clearly not working as it should. I just hope i havent wrecked my 680M.

Sent from my SM-A500FU using Tapatalk

 

If you've got a reseller near you, a test in a machine could be an option. Asking is free.

Damn orange juice, it ruins everything.

 

How long will it take you to get it replaced?

Sent from my LG-H811 using Tapatalk

 

I guess 1 week delivery time plus an evening to replace the board.

The board has seen about 6 years of life at least as many GPU's. Already 5 just with me and I already bought it 2nd hand and it has also been abused in terms of cpu OCing

Actually surprised that the power circuits failed on gpu not cpu first.

Sent from my SM-A500FU using Tapatalk