Graphics (possibly on-board) Malfunction

[adityagautam]

I have a Dell Studio XPS 1645 laptop.
(Basic configuration: Intel (first generation) Core i7-QM quad core processor @1.6 GHZ, turbo upto 2.8 GHZ, 4 GB RAM, ATI Mobility Radeon 4670 w/ 1GB DDR3)

One fine day, doing nothing particular, my screen went all crazy, with all sorts of lines on it, gave me a BSOD that was barely visible, and crashed. And crashed again and again for the next two dozen times on boot. Even the splash screen had those lines. I got some pictures to illustrate, have a look.

http://img543.imageshack.us/img543/4821/1006135f.jpg

I later viewed the contents of the BSOD, didn't convey anything in particular.

==================================================
Dump File : 031513-50372-01.dmp
Crash Time : 15-03-2013 16:49:12
Bug Check String :
Bug Check Code : 0x00000116
Parameter 1 : fffffa80`077a7010
Parameter 2 : fffff880`04a9df68
Parameter 3 : 00000000`00000000
Parameter 4 : 00000000`00000002
Caused By Driver : dxgkrnl.sys
Caused By Address : dxgkrnl.sys+5d000
File Description :
Product Name :
Company :
File Version :
Processor : x64
Crash Address : ntoskrnl.exe+75c40
Stack Address 1 :
Stack Address 2 :
Stack Address 3 :
Computer Name :
Full Path : C:\Windows\Minidump\031513-50372-01.dmp
Processors Count : 8
Major Version : 15
Minor Version : 7601
Dump File Size : 556,720
==================================================

Subsequent attempts to boot all result in a BSOD, as shown above. I did, however manage to boot in safe mode. Although, it looked like this.

http://img51.imageshack.us/img51/7659/1006131y.jpg

Also, when I disabled (and also uninstalled) my ATI graphics card drivers, I could boot normally, but with the crazy lines, so basically, the graphics card is what was causing the BSOD, although the issue still persisted, indicating the cause was something different, possibly the on board graphics processor.

Also, I really don't think it's a software issue, since even at the splash screen, the problem persisted.

http://img203.imageshack.us/img203/2513/1006138o.jpg
http://img692.imageshack.us/img692/6572/1006137zbd.jpg
http://img138.imageshack.us/img138/913/1006134r.jpg

The second image above, the appears when the screen ought to have been blank.

And another peculiar fact. After running the diagnostics tool, with the weird lines, on the next boot up, at least from sometime the problem seems to have disappeared. Although, the diagnostics screen looks like this:

http://img526.imageshack.us/img526/7859/1006141s.jpg

(Notice the weird characters at the bottom right of the screen)

This problem appeared the first time about a week ago. I did some massive graphics drivers (that time only the ATI one) uninstalling/reinstalling etc. By the time I did this, and ran some integrity check, my computer just powered off during the reboot, but that was because it was really really hot. I let it cool down, and on the next boot it booted just fine, I reinstalled all the drivers, and it ran fine for about a week. And then today it happened again. As I said, after the diagnostics, it runs fine for a while, as it is now, when I'm typing this in. Could anybody tell what the problem might be, and what could be done? Any help would be greatly appreciated.

P.S. The on-board graphics cannot be disabled, at least not from the BIOS. No such options exist.

 

[bucknutty]

so the problem happens with the onboard gpu or the AMD card. Keep in mind that the onboard GPU is built into the i7. It is not on the motherboard.

The CPU might be overheating. are all the fans spinning. in safe mode can you run coretemp? Its a free portable program that shows the temp of each core. I think this is unlikely because it would take a bit to got hot, I would think it would post and boot correctly but then overheat when in windows.

The first thing I would do is try another monitor, or at least a new cable.
Then I would test the ram. If you can get to safe mode type "memory" into the search box. That will open the memtest option. A ram stick that is messing up can cause all kinds of anomalies.

 

[adityagautam]

Hello Bucknutty,

Thanks for taking out your time to look into my problem. I have a first generation core i7, the onboard GPU didn't come built into the processor back then (or so I think). Also, I do not clearly get what you mean by try a new cable; this is a laptop computer. I ran several checks, diagnostics etc, all reports came in fine. I know what a displaced RAM stick can do, the single RAM stick on my desktop once got dislodged, and there was no signal onto the monitor. I'm certain that isn't the issue here. What bothers me is that it's an ON-OFF issue. Now it's here, now it's not. I'm not even able to zero in on the issue or the component that's causing the trouble, let alone the cause.

For the record though, I use an external program called Speedfan to check my processor temperature. Sometime back, when I was experiencing those issues, it was running at 70-75 degrees Celsius, for a mere 0-5% CPU load. This is usually a hot running computer, but that was hot even my these standards. It's normal temperature is around 60 degrees. Right now, it's running at about 57 degrees. That is the problem. It works fine, sometimes, sometimes it conks out. Do you know any utility that can diagnose my onboard graphics unit?

 

[bucknutty]

I meant to say try an external monitor, thinking of too many things at once. If the issue is with the actual decoder part of the panel, or the cable then an external monitor should work correctly. The panel and cable can get damaged, because they are moved every time you open or close the laptop.

If the problem persists with the external monitor then we know its the board, or chip. On this laptop the AMD gp is part of the mother board, it is not removable.

Your right the early i7s did not have a graphics core. I did not realize that.

I would still run memtest. It will only take a few minutes and is easy to do.

Speed fan is good. Same idea as coretemp. I recommend coretemp only because it is simple.

External monitor will rule out the panel
Mem test will rule the ram
Only thing left would be motherboard or CPU, I don't know of a good way to diagnose those.

 

[bucknutty]

Dell publishes great guides on how to disassemble their computers. Check this out. Look at how the monitor cable is routed through the hinge. That cable may have worn out.

ftp://ftp.dell.com/Manuals/all-products/esuprt_laptop/esuprt_studioxps_laptop/studio-xps-1645_service%20manual_en-us.pdf

 

[kimofrancesca]

Long overdue update - after digging through every forum I could find, I decided that my computer was pretty much a doorstop, so I didn't have much to lose if I tried something radical. So...I did! I cooked my motherboard - according to a video on YouTube - yep...cooked it! Only instead of the recommended 8 minutes, I didn't turn off the oven at the appropriate time (as I thought I had) and actually cooked the board for 20 minutes....expecting it to be a melted, burned mess. Decided to let it cool and see if by some slim chance I hadn't completely destroyed it. Well...long story short, I'm typing on the same computer at this moment - and it's been running perfectly and much cooler than it ever has. In all my researching, I came across a number of folks who indicated that Dell seemed to use a very low-quality thermal compound on their computers - even the higher-end ones. So after cooking the board and before attempting to restart, I scraped off and reapplied new compound on the gpu chip. This laptop has NEVER run so cool and I've not had a seconds trouble with it since. Likely won't last forever, but it certainly has surprised me that it's been running fine for almost 2 years beyond the original disaster. Go figure.

I have the same computer - XPS 1645 with first gen i7 - just yesterday started getting the exact same crazy screens you pictured. Have you come up with a solution? I have been unable to get laptop to boot in any mode or run diagnostics. I was able to connect to external monitor and get same screens - based on this, I assumed this had something to do with graphics card/processor, or perhaps motherboard. Interestingly, I JUST installed SpeedFan (have used in the past with Win XP Pro) immediately prior to this issue occurring. Am wondering now if SpeedFan install may have caused fault? Any suggestions or remedy would be GREATLY appreciated! 8^/

I have a Dell Studio XPS 1645 laptop.
(Basic configuration: Intel (first generation) Core i7-QM quad core processor @1.6 GHZ, turbo upto 2.8 GHZ, 4 GB RAM, ATI Mobility Radeon 4670 w/ 1GB DDR3)

One fine day, doing nothing particular, my screen went all crazy, with all sorts of lines on it, gave me a BSOD that was barely visible, and crashed. And crashed again and again for the next two dozen times on boot. Even the splash screen had those lines. I got some pictures to illustrate, have a look.

http://img543.imageshack.us/img543/4821/1006135f.jpg

I later viewed the contents of the BSOD, didn't convey anything in particular.

==================================================
Dump File : 031513-50372-01.dmp
Crash Time : 15-03-2013 16:49:12
Bug Check String :
Bug Check Code : 0x00000116
Parameter 1 : fffffa80`077a7010
Parameter 2 : fffff880`04a9df68
Parameter 3 : 00000000`00000000
Parameter 4 : 00000000`00000002
Caused By Driver : dxgkrnl.sys
Caused By Address : dxgkrnl.sys+5d000
File Description :
Product Name :
Company :
File Version :
Processor : x64
Crash Address : ntoskrnl.exe+75c40
Stack Address 1 :
Stack Address 2 :
Stack Address 3 :
Computer Name :
Full Path : C:\Windows\Minidump\031513-50372-01.dmp
Processors Count : 8
Major Version : 15
Minor Version : 7601
Dump File Size : 556,720
==================================================

Subsequent attempts to boot all result in a BSOD, as shown above. I did, however manage to boot in safe mode. Although, it looked like this.

http://img51.imageshack.us/img51/7659/1006131y.jpg

Also, when I disabled (and also uninstalled) my ATI graphics card drivers, I could boot normally, but with the crazy lines, so basically, the graphics card is what was causing the BSOD, although the issue still persisted, indicating the cause was something different, possibly the on board graphics processor.

Also, I really don't think it's a software issue, since even at the splash screen, the problem persisted.

http://img203.imageshack.us/img203/2513/1006138o.jpg
http://img692.imageshack.us/img692/6572/1006137zbd.jpg
http://img138.imageshack.us/img138/913/1006134r.jpg

The second image above, the appears when the screen ought to have been blank.

And another peculiar fact. After running the diagnostics tool, with the weird lines, on the next boot up, at least from sometime the problem seems to have disappeared. Although, the diagnostics screen looks like this:

http://img526.imageshack.us/img526/7859/1006141s.jpg

(Notice the weird characters at the bottom right of the screen)

This problem appeared the first time about a week ago. I did some massive graphics drivers (that time only the ATI one) uninstalling/reinstalling etc. By the time I did this, and ran some integrity check, my computer just powered off during the reboot, but that was because it was really really hot. I let it cool down, and on the next boot it booted just fine, I reinstalled all the drivers, and it ran fine for about a week. And then today it happened again. As I said, after the diagnostics, it runs fine for a while, as it is now, when I'm typing this in. Could anybody tell what the problem might be, and what could be done? Any help would be greatly appreciated.

P.S. The on-board graphics cannot be disabled, at least not from the BIOS. No such options exist.