P750DM-G died and won't turn on

My P750DM-G turned off while I was away from the desk and now it won't turn on

I suspect the GPU may be cooked. It's the only thing I can think of why this would occur. Because recently I set the NVIDIA performance mode to "Prefer Maximum". I typically only turn this on for gaming but forgot to set it back to "Auto" after gaming a few nights ago.

I also noticed over the past couple days HD playback on YouTube has been choppy. I didn't think anything of it really, but now am linking this to the current situation and seems like that could be an indication the GPU has been running at crazy temps.

Certainly a driver setting shouldn't cause this, but if the GPU was flaky then maybe it was already on it's way out.

Of course, I am open to other ideas. Still under warranty but I'd like to fix this here if possible to save sending it for repair.

If you have some specific suggestion or if there is a recovery technique with these Clevo units I would be willing to try. Thanks.

 

Does the computer have any lights come on when you try to start it up? Does it make any noises or the fans start up? Have you checked to make sure that everything internally is seated properly? what all have you done so far to try to troubleshoot?

 

Without AC plugged in, pressing power results in the power indicator light to briefly flash green once.

With AC plugged in, sitting idle the green HDD light is a solid green and the power light is a solid yellow.
Pressing power will turn off both lights briefly, followed by the HDD light turning solid yellow for a few seconds, then turns back to solid green. The power light remains a solid yellow. No other lights turn on and no fans or anything else turns on; totally silent.

I tried removing the battery and also powering it on with just AC with no results.
And I removed the two HDD's that were in there to begin the process of file recovery.

The system has never been opened after it's original build only a few months ago so I can't imagine there would be any seating issues - also it was stationary for about 2 days until it suddenly died today. But if there's anything specific you would suggest to check do share. Thanks.

 

It's certainly not because of Maximum Performance mode... Hardware should be stable with it even 24h until warranty end (but if cooled properly and GPU, but we have throttling). It might just "release the Kraken", i mean GPU was faulty and it showed that earlier (good, better now than in future).

Did you tried to enter Windows in SAFE mode [with internet]? Uninstall nVidia drivers and install older version, for example 359.06 or 362.00.
Have you installed 364.XX drivers? They are faulty and shouldn't be installed! Even 364.51 WQHL overheats GPU (confirmed).

 

Did you try removing the bios battery and starting it up?

As in removing it for couple of minutes to reset the ec, putting it back in and then starting up the lappie.

Sent from my Nexus 5 using Tapatalk

 

Oh interesting, just so happens that a few days ago I upgraded from 352.63 to 361.28 (Linux)
Is 361.28 considered problematic for Clevo's ??

 

It's just the 364. versions that removed from download by nvidia.

 

After removing the CMOS battery / holding power for 30 seconds, and then plugging it in (without large battery) - the lights flashed green briefly - and then nothing; no lights and totally dead; unresponsive to power button.

After putting back in the large battery pressing power the both lights flash and go solid yellow briefly - then the battery light goes solid green and power light remains solid yelow; but still no fans turn on / total silence.

Then same results after putting the CMOS battery back in.

 

Looks like there are still some Linux users reporting an issue with 361.28 indicating that a previously fixed bug "has reappeared in 361.28".

https://devtalk.nvidia.com/default/...on-close-only-with-proprietary-nvidia-driver/

The issue is 100% CPU usage of one CPU core (for a process that runs in the background until you kill it.... or until it cooks that core i suppose?). It doesn't seem like my system has that exact config as those users, for example I'm running Ubuntu not Kubuntu, but I do have a similar config and the issue does sound telling.

So my next question is, what about the automatic shutdown? If dangerous temps are reached does not this Clevo power down automatically? And if that is a feature, is it compatible with Linux? I was under the assumption that if the CPU or GPU reached critical temp that the system would shut down, and would do so regardless of OS.

 

Temperature sensing is OS independent and managed at a firmware level.

 

I had this exact issue.
The end result was an RMA.

I believe it could be BIOS corruption.
Typically on Clevos there is a recovery mode that may help. Unfortunately I don't have the files to share with you and even when I did I could never get it working. Usually you'd use a USB 2.0 port but since there are none...

Hold FN+R after CMOS reset and if you enter some kind of recovery I may be able to dig out those files or @Prema might be nice enough to share but ultimately I do believe you'll be looking to RMA.

If you have a flash drive that indicates power do you find that it only responds in the eSata port?
It was the only oddity I found.

 

https://www.techinferno.com/index.p...ods-prema-mod™-stock/&page=150#comment-140641

 

You can also ask your reseller for a stock file if you want to keep the splash you currently have on start up.

 

Thanks for sharing your similar experience @stonetrap
I also did an RMA. Hopefully the issue can be pinpointed to avoid the same thing happening again.

 

Still waiting on my laptop from RMA so I don't have any specifics about what the problem is exactly but maybe it is a dead CPU due to Skylake + Linux issues:

https://news.ycombinator.com/item?id=11492070
http://mjg59.dreamwidth.org/41713.html

@stonetrap were you running Linux ? And either way, were you able to find out the root cause?

 

From what I can tell regarding those posts (also don't know if this was mobile Skylake vs. Desktop), it looks like the problem is one wouldn't get very good battery performance when running Linux.

 

Performance and battery life per charge is not great on a mobile workstation/desktop replacement, irrespective of the OS. (The P750ZM/DM uses a desktop chip.)

 

That wasn't my point.

What I was getting at was the previous post about dead CPU maybe attributed to running Linux ("maybe it is a dead CPU due to Skylake + Linux issues"), but the links for the Linux posts were not about killing the CPU but rather that it had diminished battery time.

 

I would recommend re-seating your memory and CMOS battery for good measure. If you have a spare CMOS battery, definitely replace it. Most of the time a device will always boot with a dead CMOS battery, but in my field I occasionally see some devices not properly power on because the dead CMOS battery tricks the mainboard in to thinking there's no power at all.

**These are do at your own risk and only recommended if you feel comfortable doing so!**

The next step to try is to verify that its not hardware related, remove each component 1 by 1 starting with your HDD and then memory to see if you can get it to power on. If any component shorts out, most of the newer boards now will not power on until its removed, the fact that your power light flashes each time you reset your CMOS tells me your board is most likely detecting a short. I've see wireless USB mouse dongles left connected sometimes prevent a device from powering on, mostly in desktop though.

After the HDD, you'll want to try disconnecting your keyboard and memory. If it still does not power on, you need to remove GPUs, and if you really want to go as far as removing your CPU and try to power it on, your BIOS will at least tell your no CPU detected. If none of this works, its A) your motherboard is cooked, B) your PSU is done, or C) broken solder contacts somewhere on the board.

Good luck!