Hello there!
I need some help with my laptop GPU.
I have a 2012 Sager NP9150 with a GTX 675m.
A couple of months ago the computer started acting up. I would be in-game and it would suddenly freeze, hang, the status LED's would flash about 20 times while the computer beeped, the fans would hit turbo, and the computer would shut off.
This has been happening for months. I've taken the following steps to try and fix it:
I've also done CPU and GPU stress tests.
- cleared all dust
- re-applied thermal paste to CPU and GPU
- Wiped and then re-installed drivers
- Wiped and clean-installed Windows 10
- Put graphics way down to low (can usually run on high)
- Put the computer on high performance mode
The CPU and GPU both reached 100C and were still fine for more than 15 minutes.
I appreciate any help and suggestions!
-
It sounds like the GPU is dying despite it passing your stress test.
Also on a side note, you should never let a GPU run as hot as 100C.bennyg likes this. -
That's what I was starting to believe as well.
Not to worry, I never have the GPU or CPU run anywhere near 100C, they are usually contained to the 80's under load (100C was just a one-off for the stress test).
Besides taking it on the chin, can anything be done for the GPU to recover some performance or should the computer just remain locked to the iGPU? -
Damn... sounds like you cleared all the moisture out of the unit at least
And I agree. I had a GPU do the slow death as well -
Silver lining of sorts I guess.
Did your machine just turn into a straight up productivity station then at that point? -
Yup. My other option at the time was to try and revamp an old an old Inspiron... which, years later, is still floating around here somewhere with its original massive allotment of 2 x 2GB of dual channel RAM still intact
lol -
What I still don't understand is why the GPU was fine during stress testing but shut the computer down while in-game for 5 or 10 minutes.
I've also heard the PSU get mentioned as a possible source of troubles, is that feasible at all?Last edited: Nov 25, 2017KY_BULLET likes this. -
Could be the PSU. Sounds a little like this case here. Probably be the cheapest fix you could try first.
http://forum.notebookreview.com/thr...be-dead-recommendations.811047/#post-10636937 -
A bad PSU doesn't cause crashes. I'd only suspect PSU if you had hard instant shutdowns.
-
Here's another piece of information I wanted to mention: every time the crashes happened, I would check out event viewer and it would show a Kernel Power Error ID - 41.
Here's the relevant site: https://support.microsoft.com/en-us...1-error-the-system-has-rebooted-without-clean
Now 41 narrowed it down to OC, overheat, RAM, defaults, or PSU. I haven't OC'd, heat is not an issue, RAM checks came up clean, defaults are fine because of clean install, so that only left the PSU for me.
What are your thoughts about this? -
Stress tests usually hit the core hard but don't stress memory as much, so don't always show up memory errors.
My brothers P150HM with 580M (same GPU as you, yay for rebadging chicanery) had exactly these kind of intermittent shutdowns, I cleaned everything out and repasted so it was no longer thermal throttling, then two weeks later it was artifacting with green squares everywhere, still stable apart from that. Gone with new gpu. -
That makes sense, good to hear that you got the HM working again.
At this point though, I don't know if it's worth it to pay ~$200 for a used GPU that is 5 years old already or if I should invest that money into something else.
Did you get your replacement off of ebay? If so, the only options I see on there are pre-owned and used cards. -
Sometimes it's people selling a broken laptop and you can part out and resell the bits you don't need (e.g. I got a complete spare P370EM shell, all working except GPU, with XM CPU in it $400 when the cheapest XMs on ebay were $250, but since I wanted to mod stuff havings spares of everything on hand was what I was after)
[Sager NP9150] dGPU Issues
Discussion in 'Sager and Clevo' started by fleeon, Nov 25, 2017.