NVIDIA Driver 388.13 Crashing/Recovering...

Message boards : Number crunching : NVIDIA Driver 388.13 Crashing/Recovering...
Message board moderation

To post messages, you must log in.

1 · 2 · Next

AuthorMessage
Profile TimeLord04
Volunteer tester
Avatar

Send message
Joined: 9 Mar 06
Posts: 21140
Credit: 33,933,039
RAC: 23
United States
Message 1955038 - Posted: 12 Sep 2018, 19:44:03 UTC
Last modified: 12 Sep 2018, 19:45:16 UTC

I keep having Video Driver Crashes under 388.13 on Prometheus, (i7 7700K System), with Dual EVGA GTX-1050s...

I just reinstalled the Driver CLEAN and it happened AGAIN!!! ALWAYS while surfing on FF Quantum, (current Version), changing Tabs, Reloading a Tab, or scanning through a Page, (scrolling), THIS IS DRIVING ME BATTY!!!

I've been lucky to date, and this is the ONLY System I've had NVIDIA Driver trouble with in Windows. (Win7 Pro x64 SP-1.)

ANY suggestions are welcome.

(The NVIDIA Drivers were Upgraded from 353.30. The 353.30 WHQL Driver was controlling EVGA GTX-750TI SC Cards, upon Upgrading to the GTX-1050s I Clean Installed 388.13 WHQL.)


TL

[EDIT:]

Driver DOES NOT crash if System is left alone, AND DOES Crunch! NO crashes while Crunching...


TL
TimeLord04
Have TARDIS, will travel...
Come along K-9!
Join Calm Chaos
ID: 1955038 · Report as offensive
Ian&Steve C.
Avatar

Send message
Joined: 28 Sep 99
Posts: 4267
Credit: 1,282,604,591
RAC: 6,640
United States
Message 1955040 - Posted: 12 Sep 2018, 20:15:05 UTC - in response to Message 1955038.  
Last modified: 12 Sep 2018, 20:18:30 UTC

use DDU to wipe out the old drivers in safe mode, and clean install the newest drivers 399.24


download DDU here: https://www.guru3d.com/files-details/display-driver-uninstaller-download.html
reboot the system into safe mode.
use DDU in safe mode to erase all traces of your previous drivers.
reboot back into normal mode.
download latest drivers version 399.24 from nvidia.com
install drivers WITHOUT geforce experience (it's trash)

if you're still having issues after this, look into maybe a hardware problem, MAYBE even some level of OS corruption (but probably unlikely)
Seti@Home classic workunits: 29,492 CPU time: 134,419 hours

ID: 1955040 · Report as offensive
Profile Cliff Harding
Volunteer tester
Avatar

Send message
Joined: 18 Aug 99
Posts: 1432
Credit: 110,967,840
RAC: 67
United States
Message 1955052 - Posted: 12 Sep 2018, 20:58:28 UTC

Check for those dastardly dust bunnies. Over the weekend my twin EVGA GTX 750Ti on my i7/7700K was crashing and bringing down the system, but only when crunching data. I blew out the dust bunnies and even took a tooth pick to clean out between the cooling vanes. When the system came back up, temps dropped 10c. BTW, I'm using the 391.01 driver.


I don't buy computers, I build them!!
ID: 1955052 · Report as offensive
Profile Jord
Volunteer tester
Avatar

Send message
Joined: 9 Jun 99
Posts: 15184
Credit: 4,362,181
RAC: 3
Netherlands
Message 1955058 - Posted: 12 Sep 2018, 21:43:50 UTC
Last modified: 12 Sep 2018, 21:45:52 UTC

I think it's an Nvidia thing. My TV server ran for years with an AMD HD7870 until it blew itself up. But all those years not a single problem with the drivers. I could reboot, hibernate or leave the system on for hours. After the AMD we put in an Nvidia GTX 1050 and since that time, about every time the computer comes out of hibernation or has had a reboot, the driver crashes. Guess you just shrug about it. We do.

(Before the DDU advice comes, prior to installing the Nvidia drivers, I ran the AMD driver uninstall program, I ran DDU, I ran one other program just to be sure. There are no remnants of AMD drivers on that system, afaik.)
ID: 1955058 · Report as offensive
Profile TimeLord04
Volunteer tester
Avatar

Send message
Joined: 9 Mar 06
Posts: 21140
Credit: 33,933,039
RAC: 23
United States
Message 1955078 - Posted: 12 Sep 2018, 23:22:04 UTC - in response to Message 1955052.  

Check for those dastardly dust bunnies. Over the weekend my twin EVGA GTX 750Ti on my i7/7700K was crashing and bringing down the system, but only when crunching data. I blew out the dust bunnies and even took a tooth pick to clean out between the cooling vanes. When the system came back up, temps dropped 10c. BTW, I'm using the 391.01 driver.

NO Dust Bunnies here... The System, (NOT including the Hard Drive), is approx. two months old... The Hard Drive is 4 Years Old and came from Hackintosh-Andromeda/Original Prometheus. My Computer Hardware Source reconditioned the Drive to boot and work with the new Hardware WITHOUT the need to reinstall the OS from Scratch... All my Data, my Programs, BOINC, are all Original Prometheus.

I will be running DDU, as described a couple posts back... I was hoping to NOT have to do that; BUT, I guess I really should have done it to remove all of 353.30 to TRULY install 388.13 Clean...


TL
TimeLord04
Have TARDIS, will travel...
Come along K-9!
Join Calm Chaos
ID: 1955078 · Report as offensive
Profile TimeLord04
Volunteer tester
Avatar

Send message
Joined: 9 Mar 06
Posts: 21140
Credit: 33,933,039
RAC: 23
United States
Message 1955082 - Posted: 12 Sep 2018, 23:55:51 UTC - in response to Message 1955040.  

use DDU to wipe out the old drivers in safe mode, and clean install the newest drivers 399.24


download DDU here: https://www.guru3d.com/files-details/display-driver-uninstaller-download.html
reboot the system into safe mode.
use DDU in safe mode to erase all traces of your previous drivers.
reboot back into normal mode.
download latest drivers version 399.24 from nvidia.com
install drivers WITHOUT geforce experience (it's trash)

if you're still having issues after this, look into maybe a hardware problem, MAYBE even some level of OS corruption (but probably unlikely)

I was hesitant to grab the Latest Driver... I'm NOT one for "Cutting Edge"... Though I thank you for your instructions. :-)

Ran DDU in Safe Mode, rebooted, reinstalled 388.13 Clean and without GeForce Experience. Rebooted AGAIN, brought back up BOINC and FF. BOINC shows CUDA and OpenCL Installed and Operating. (As they were before...)

I Reloaded EVERY Tab I had from my Prior FF Session, Logged into Amazon and compared specs on a couple of MOBOs, (which is where the Driver CRASHED last time), and everything seems OK now. I will continue to monitor the system, but I think it's fine, now. :-)


TL
TimeLord04
Have TARDIS, will travel...
Come along K-9!
Join Calm Chaos
ID: 1955082 · Report as offensive
Profile TimeLord04
Volunteer tester
Avatar

Send message
Joined: 9 Mar 06
Posts: 21140
Credit: 33,933,039
RAC: 23
United States
Message 1955088 - Posted: 13 Sep 2018, 0:15:47 UTC

Just played a YouTube Video - NO Glitching, NO Driver Crash... :-) Then went to Plex, played a DS-9 Episode for a few minutes - NO Glitching, NO Driver Crash... :-)

I think, (and HOPE AND PRAY), that everything is as it should be. :-) :-D


TL
TimeLord04
Have TARDIS, will travel...
Come along K-9!
Join Calm Chaos
ID: 1955088 · Report as offensive
Ian&Steve C.
Avatar

Send message
Joined: 28 Sep 99
Posts: 4267
Credit: 1,282,604,591
RAC: 6,640
United States
Message 1955097 - Posted: 13 Sep 2018, 1:00:24 UTC

good to hear it's fixed. the big red flag for me with your issue was that you upgraded from an older version to a newer version and it sounded like it installed over top one another. most cases that's ok, but can occasionally cause driver issues like that.

its ok to take a "if it aint broke, dont fix it" stance for nvidia drivers, but for me personally i have never had a problem with WHQL drivers (399.24 is WHQL) and i occasionally update them (always wiping out the old before install). but i would agree with you about bleeding edge BETA drivers. no need to use them unless you want/need to test a new feature.
Seti@Home classic workunits: 29,492 CPU time: 134,419 hours

ID: 1955097 · Report as offensive
Profile TimeLord04
Volunteer tester
Avatar

Send message
Joined: 9 Mar 06
Posts: 21140
Credit: 33,933,039
RAC: 23
United States
Message 1955110 - Posted: 13 Sep 2018, 2:19:33 UTC - in response to Message 1955097.  

good to hear it's fixed. the big red flag for me with your issue was that you upgraded from an older version to a newer version and it sounded like it installed over top one another. most cases that's ok, but can occasionally cause driver issues like that.

its ok to take a "if it aint broke, dont fix it" stance for nvidia drivers, but for me personally i have never had a problem with WHQL drivers (399.24 is WHQL) and i occasionally update them (always wiping out the old before install). but i would agree with you about bleeding edge BETA drivers. no need to use them unless you want/need to test a new feature.

I DO like Close To Cutting Edge on Hardware, but, yeah - Software and Drivers I tend to hold back from Bleeding/Cutting Edge. 4 Years ago, when Original Prometheus was first made, I started with 266.58 for a Pre-Fermi GTX-275, then when I got my first GTX-750TI SC, (used), from Zalster I Upgraded to 337.88 and stayed there for most of Prometheus' life. After reading multiple Posts here about 353.30 when many people here were using that, I decided to make the move to it. THEN, I found out about the Hackintosh Scene and got a second Drive and turned the System into Hackintosh-Andromeda/Prometheus-Win7 Pro and remained on the MacOS side for the past two+ years...

Hackintosh, (as Posted in my other Thread), died a couple months ago and I thought the newer Hardware I had on hand was 100% Hackintosh Compatible - BUT found that the Gigabyte GA-Z270-HD3 MOBO needs tweaking in the USB Flash Drive Installer to get to the MacOS Installer... NOT quite as friendly as my old Gigabyte GA-EP45-UD3P was...

Did find; however, that the GA-H270-HD3 IS FULLY 100% Hackintosh Compatible, made one for a friend, made another for a friend-of-a-friend... So, just picked up that board for myself, and in the next 6Mos. hope to be able to get enough Credit Card Debt paid down to get RAM and a PSU. (Getting a Case from Carlos, soon.) So, Hackintosh-Andromeda WILL be back. For now, I changed the MacPro 5,1 name from Daedalus to Andromeda and moved the Hackintosh-Andromeda BOINC Data Folder over to it...

Anyway, New Prometheus on the i7 7700K seems pretty fast and powerful. MOST powerful PC I've EVER had! I even got USB 3 Drivers from Gigabyte for the Z270 MOBO for Win7 Pro!!! I won't, (thankfully), have to deal with Win 10 or whatever replaces that for several years to come, now. I'm also hoping that in the next 6Mos. that I can afford to pick up a Samsung 850 or 860 Pro to Clone New Prometheus over to... Especially since the WD Black 1TB Drive is now in it's 4th Year of Life... (I'll do the Cloning before doing anything more with the Hackintosh...)


TL
TimeLord04
Have TARDIS, will travel...
Come along K-9!
Join Calm Chaos
ID: 1955110 · Report as offensive
Profile TimeLord04
Volunteer tester
Avatar

Send message
Joined: 9 Mar 06
Posts: 21140
Credit: 33,933,039
RAC: 23
United States
Message 1955270 - Posted: 13 Sep 2018, 22:33:54 UTC

Well, IT HAPPENED AGAIN!!!!! :-O

Was pricing parts via Amazon.com for a computer build for a friend, and comparing two different MOBOs the NVIDIA Driver FAILED!!! Screen went black for 20 Seconds this time!!!


TL
TimeLord04
Have TARDIS, will travel...
Come along K-9!
Join Calm Chaos
ID: 1955270 · Report as offensive
Ian&Steve C.
Avatar

Send message
Joined: 28 Sep 99
Posts: 4267
Credit: 1,282,604,591
RAC: 6,640
United States
Message 1955288 - Posted: 14 Sep 2018, 1:09:32 UTC

cant hurt to try a new driver version. just in case.
Seti@Home classic workunits: 29,492 CPU time: 134,419 hours

ID: 1955288 · Report as offensive
Profile Zalster Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 27 May 99
Posts: 5517
Credit: 528,817,460
RAC: 242
United States
Message 1955289 - Posted: 14 Sep 2018, 1:18:46 UTC - in response to Message 1955288.  

I never upgrade there's a real reason to do so...
ID: 1955289 · Report as offensive
Grant (SSSF)
Volunteer tester

Send message
Joined: 19 Aug 99
Posts: 13720
Credit: 208,696,464
RAC: 304
Australia
Message 1955310 - Posted: 14 Sep 2018, 6:24:51 UTC - in response to Message 1955288.  

cant hurt to try a new driver version. just in case.

If it isn't broken, don't fix it.

The huge majority of driver updates are usually to support, or improve support, for games. There are no changes to CUDA or OpenCL performance.
Then you'll get new drivers that support new hardware, so you'll need those drivers if you've got that new hardware. And unless they really screwed the compute side of them up, none of the following driver releases will have any effect on compute times.
Grant
Darwin NT
ID: 1955310 · Report as offensive
Ian&Steve C.
Avatar

Send message
Joined: 28 Sep 99
Posts: 4267
Credit: 1,282,604,591
RAC: 6,640
United States
Message 1955332 - Posted: 14 Sep 2018, 14:28:35 UTC
Last modified: 14 Sep 2018, 14:48:56 UTC

but his system is broken in some way. he keeps getting driver crashes. and as TL stated, the problems appeared to start with installing this 388 version.

trying a new driver is the easiest first step. i don't care if he doesn't use 399 or even reverts to an older version, just use a different driver. eliminate that as a problem, move on to the next possible cause. if the driver keeps crashing with a different version, then move on to the possibility of a hardware issue, or some amount of OS or conflicting software issues. basic troubleshooting.
Seti@Home classic workunits: 29,492 CPU time: 134,419 hours

ID: 1955332 · Report as offensive
Ian&Steve C.
Avatar

Send message
Joined: 28 Sep 99
Posts: 4267
Credit: 1,282,604,591
RAC: 6,640
United States
Message 1955335 - Posted: 14 Sep 2018, 14:42:36 UTC
Last modified: 14 Sep 2018, 14:52:14 UTC

another thought. TL stated no crashes while crunching alone.

does the driver crash while Browsing + Crunching?
does the driver crash while Browsing and NOT crunching?

if the former, what kind of cmdline settings are you running for your GPU jobs? maybe it's too aggressive for the 1050 to handle browing and crunching at the same time.
if the later, does it crash with other browsers? chrome? or IE? maybe try a full uninstall/reinstall of firefox
Seti@Home classic workunits: 29,492 CPU time: 134,419 hours

ID: 1955335 · Report as offensive
Profile TimeLord04
Volunteer tester
Avatar

Send message
Joined: 9 Mar 06
Posts: 21140
Credit: 33,933,039
RAC: 23
United States
Message 1955336 - Posted: 14 Sep 2018, 15:17:03 UTC
Last modified: 14 Sep 2018, 15:23:51 UTC

Just for clarity... FF is always up/launched and has 25-35 Tabs Open at a time... When I say I'm "surfing" or using FF, it's when I'm sitting at the computer; the computer runs 24/7 BUT ONLY crunches from 6PM to 9AM Pacific. So, to answer one of the questions, Yes - FF is on while crunching...

Now, maybe this has been happening even while I'm away from the System and, thus, I haven't been able to witness it until recently... Or, it's something new just happening, and ONLY happening when I'm fiddling around with FF...

Another thing, is that this System, (as it stands, now, with the new Hardware), has been up for about a couple months. One NEW change is that I've moved it from SETI crunching to Collatz crunching over the past couple of weeks. (Since the end of WOW.) About a week ago, I "Optimized" Collatz per instructions at that Site. The Line "sieve_size=30" was a Default Parameter, and over the course of a couple evenings of crunching, when I was happening to sit at the System to make use of it WHILE crunching, I noticed Sluggishness in Mouse movement and in switching Tabs in FF.

Found that I had to Reduce the "sieve_size=30" to "sieve_size=27" to stop the sluggishness while trying to make use of the System... COULD Collatz, (somehow), have damaged Dev 0, (the 1050 connected to the monitor), or do we STILL believe this is just a Driver Issue??? This Driver Failure seemed to Pop-up in the last week - THAT I've noticed... However; IF this IS just a Driver Issue, it could, (as suggested), have been going on since installing 388.13 two months ago when I installed the first of the 1050s into the System...

Food for thought...


TL

[EDIT:]

[Collatz Optimization Settings:]

[EVGA GTX-1050 2GB GDDR5 VRAM Optimized Settings - SWEET SPOT:]

verbose=1
kernels_per_reduction=48
threads=9
lut_size=17
sieve_size=27
reduce_cpu=0
cache_sieve=1
sleep=0
TimeLord04
Have TARDIS, will travel...
Come along K-9!
Join Calm Chaos
ID: 1955336 · Report as offensive
Ian&Steve C.
Avatar

Send message
Joined: 28 Sep 99
Posts: 4267
Credit: 1,282,604,591
RAC: 6,640
United States
Message 1955343 - Posted: 14 Sep 2018, 16:08:13 UTC
Last modified: 14 Sep 2018, 16:10:31 UTC

it sounds like you have a lot of things going on with the system in general. I don't think anyone will be able to say with certainty "THIS is exactly your issue". You'll have to go through a bit of trial and error and try different fixes and see what works. changing the driver is easy, which is why i suggested it. you're having driver crashes. changing the driver itself is easy, fast, and wont make anything worse.

also, is there any particular reason you leave FF open all the time with so many tabs? just convenience so you don't have to re-open them?

does the driver ever crash between 9AM-6PM when the system is NOT crunching?

unfortunately i dont know anything about Collatz, i thought you were having issues while crunching SETI. is there a forum for collatz to discuss settings? maybe someone there or someone here who knows more about those settings could provide more input about if they are too aggressive or not. i just can't say myself. but from a seti point of view, aggressive GPU settings certainly can cause screen lag.

i would try these options. in order of increasing amount of work to do/cost.

1. try a new/different driver
2. try less aggressive compute settings
3. try a new browser (even as just a test, not a long term solution, to rule out FF causing the issue)
4. reinstall firefox.
5. swap the video cards in the system, see if the problem stays with one card driving the monitor vs another.
6. re-install your Windows OS + apps from scratch.
7. replace physical hardware (GPUs/motherboard/etc)
Seti@Home classic workunits: 29,492 CPU time: 134,419 hours

ID: 1955343 · Report as offensive
Profile tullio
Volunteer tester

Send message
Joined: 9 Apr 04
Posts: 8797
Credit: 2,930,782
RAC: 1
Italy
Message 1955348 - Posted: 14 Sep 2018, 16:28:36 UTC
Last modified: 14 Sep 2018, 16:29:31 UTC

I am not a GPU expert. I have a GTX 1050 Ti installed in my Windows 10 PC with its A10-6700 AMD CPU, not overclocked. I have the following results; Einstein@home GPU tasks run peacefully, with a low error percentage. SETI@home GPU tasks run also, but cause frequent reboots, but the tasks complete and validate,no errors. GPUGRID GPU tasks run, but the GPU reaches 80 C and then the task crashes, without causing a reboot. So I am mostly running Einstein@home tasks on it, leaving SETI@home and GPUGRID GPU tasks to my Linux box with its GTX 750 Ti which never goes above 70 C.
Tullio
ID: 1955348 · Report as offensive
Profile TimeLord04
Volunteer tester
Avatar

Send message
Joined: 9 Mar 06
Posts: 21140
Credit: 33,933,039
RAC: 23
United States
Message 1955359 - Posted: 14 Sep 2018, 18:12:37 UTC - in response to Message 1955343.  
Last modified: 14 Sep 2018, 18:15:13 UTC

it sounds like you have a lot of things going on with the system in general. I don't think anyone will be able to say with certainty "THIS is exactly your issue". You'll have to go through a bit of trial and error and try different fixes and see what works. changing the driver is easy, which is why i suggested it. you're having driver crashes. changing the driver itself is easy, fast, and wont make anything worse.

also, is there any particular reason you leave FF open all the time with so many tabs? just convenience so you don't have to re-open them?

does the driver ever crash between 9AM-6PM when the system is NOT crunching?

unfortunately i dont know anything about Collatz, i thought you were having issues while crunching SETI. is there a forum for collatz to discuss settings? maybe someone there or someone here who knows more about those settings could provide more input about if they are too aggressive or not. i just can't say myself. but from a seti point of view, aggressive GPU settings certainly can cause screen lag.

i would try these options. in order of increasing amount of work to do/cost.

1. try a new/different driver
2. try less aggressive compute settings
3. try a new browser (even as just a test, not a long term solution, to rule out FF causing the issue)
4. reinstall firefox.
5. swap the video cards in the system, see if the problem stays with one card driving the monitor vs another.
6. re-install your Windows OS + apps from scratch.
7. replace physical hardware (GPUs/motherboard/etc)



1. try a new/different driver - I'm open to this.

2. try less aggressive compute settings - The Collatz Settings listed ARE the Less Aggressive Settings, now.

3. try a new browser (even as just a test, not a long term solution, to rule out FF causing the issue) - The ONLY other Browser installed and available is IE for Win 7... NOT going there... ;-)

4. reinstall Firefox. - Can try this as a last resort.

5. swap the video cards in the system, see if the problem stays with one card driving the monitor vs another. - I was thinking this same thing after my last Post.

6. re-install your Windows OS + apps from scratch. - NOT happening! :-O *shudder*

7. replace physical hardware (GPUs/motherboard/etc) - ALL Hardware is Brand New, (except the Hard Drive - 4 Years Old), and IF you would like donate US $$$ to the cause, I'd THEN be willing to do this... Otherwise, NOT happening on my limited budget... (It took me 4-5 Months to get the Hardware I have now.)

[EDIT:]

The Driver Crashing happens between 9AM and 6PM, whenever I happen to be on the System... I try to leave the System alone when crunching between 6PM and 9AM, but occasionally am on to check e-Mail, or browse Amazon... Maybe even check Updated Posts in various Projects. So far, Driver crashing DOES NOT occur while crunching - even if I'm on the System doing something else.


TL
TimeLord04
Have TARDIS, will travel...
Come along K-9!
Join Calm Chaos
ID: 1955359 · Report as offensive
kittyman Crowdfunding Project Donor*Special Project $75 donorSpecial Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 9 Jul 00
Posts: 51468
Credit: 1,018,363,574
RAC: 1,004
United States
Message 1955360 - Posted: 14 Sep 2018, 18:24:32 UTC
Last modified: 14 Sep 2018, 18:25:01 UTC

The most obvious thing to me would be to go back to crunching Seti and see if the problem shows up.
Different computing applications can be very different in how they use the computer's resources.
And since you indicated that moving to Collatz was a recent change, that could change a lot.
"Freedom is just Chaos, with better lighting." Alan Dean Foster

ID: 1955360 · Report as offensive
1 · 2 · Next

Message boards : Number crunching : NVIDIA Driver 388.13 Crashing/Recovering...


 
©2024 University of California
 
SETI@home and Astropulse are funded by grants from the National Science Foundation, NASA, and donations from SETI@home volunteers. AstroPulse is funded in part by the NSF through grant AST-0307956.