Problem amd 7970M who helps me solve this problem?

Hello everyone I hope you can help me!
first of all sorry for my english I know that it is not good
The problem is this: when I use the gpu for editing, gaming, the fans of my clevo p370em after a while they start to turn up, the temperatures are normal between 66 and 84 degrees C
starts an audible alarm scroll anc capslock icon start lighting and after 10-20 seconds, my clevo automatically shutdown
AMD 7970M is stock, no ov VBIOS or other
clevo bios and ec firmware.are latest version
my problem is very similar to this:

HTML:

http://www.youtube.com/watch?v=_K96eNSyFkA

The difference lies in the fact that my computer boot, it starts the alarm sound when I use the video card for example graphics tests or gaming,

Clevo p370em
cpu i7 3740qm
gpu 1 x AMD 7970M 2GB DDR5
8 gb ram
intel ssd 520
scorpio black 500 gb hdd


If anyone knows how to help me I would be grateful, I do not know how to fix it!

 

Sell it.

Buy a 680M for a bit more, or get a 675MX which overclocked can overtake your 7970M. The soldering, manufacturing of the Clevo 7970M is probably among the worst I have seen on any hardware. I see better workmanship on $10 toys. Dell/MSI 7970M don't see to have the same QC disaster.
- I took a look at my 7970M when it too failed, the soldering was burnt on 3 chips and on one, it fell off. Temps were always cool, around 72C and I have a 240watt PSU, so don't think heat or power was the reason. It's just the worst soldering, QC, workmanship on any computer hardware, probably sets the benchmark for bad crap.

Shoddy driver support, Enduro still not fixed as they promised, the worst manufacturing, just kiss your AMD goodbye and get something of higher quality.
- Also talked to someone who works with GPU drivers, Nvidia drivers are often more hardware based which is why there aren't as much difference between Linux/Windows drivers like on AMD. AMD drivers are nearly all software based, which is why the hardware engineering is great, but the drivers suck so bad. And also why AMD drivers suck so much on Linux.

 

You could try to run memtest to check your ram. It might not be the 7970m causing the problem.

This sounds a bit like this scenario:

http://forum.notebookreview.com/sager-clevo/715644-amd-7970m-failing-no-overheat.html

 

thank you very much for your response.
Well I can not sell a gpu probably broken
I agree with everything you said,
I keep most definitely consider nvidia, compared to amd
I would like to understand if the problem is just the gpu
if I could solve, probably sell it anyway

 

ok thanks for the advice I try to test the ram will let you know.
thanks for advice guys

 

Yup, it is quite possible (I'm the OP of that refered Post). I was quick to blame the GPU and it wasn't faulty, a RAM module was. Try running memtest, as the above poster said, and try removing some or moving around your RAM modules just to be safe. Check if your GPU and CPU are correctly seated as well.

 

hi using this tool memtestcl for vram testing reported 32 errors how to restore a VRAM video card?

screenshot: http://s21.postimg.org/hd6w73vuv/memtestcl.jpg

 

Quite odd, the errors where only with random blocks.

I'd still go with possible RAM issues. Why?, Well, the test will write random blocks on the VRAM and will keep a reference of those same written blocks in RAM for comparing when finished. So, if they don't match, it will say that the VRAM is at fault, but it can really as well be the RAM no being capable of holding those values correctly. The sequential test with 1s and 0s showed no error (those tests rely on filling VRAM with 0s and checking if there are 1s and vice versa) uses way less RAM. I never seen a defective VRAM pass random 1s and 0s...

Have you tested your ram as previously suggested?

If it is indeed VRAM, you are pretty much screwed. I think there was a tool to limit the usage of VRAM on a board, but if the damaged addresses aren't sequential, it will probably do no good. If so, prepare to shell out :S

 

I'm already preparing to shell out money!
I have also tested the ram, I run memtest for an hour, until the test module 20,
not scored errors, I have to run the test for more time?
gpu and cpu are correctly seated

 

Run it overnight and see how you go.

 

Hi Columbo

ok done!
memtest did 3 reps in the memory test, no problem detected.
it seems the problem is not in the ram.
I do not want to buy a brand new gpu and maybe discover that the problem was not the graphics card, I can not understand,
why the problem occurs only when I stress the gpu?
the temps are good,no throttling,at a certain moment, I hear the alarm sound, the fans runs at maximum spin,after 20 seconds shutdown without bsod or any errors
this does not make sense!

 

Can't you RMA the laptop? What kind of alarm sound do you hear? Beeps? How many? Short or long?

 

is no longer under warranty
beeping alarm
caps lock and ScrLk, start flashing as when you update the bios of the laptop
the fan runs at maximum, after 8-10 beep shutdown the laptop.
as if he had temperature problems, but they are normal

 

Computer POST and beep codes

 

may seem like a stupid suggestion but have u tried cleaning it up from inside so that dust isnt clogging any airflow?

 

Or have you checked the fans are operating as they should (despite the temps)?

 

no you were right to write it!
blown mobo gpu and the whole interior of the laptop
repasted cpu and gpu
checking the fans and they work
updated bios and ec firmware
at last version
Check RAM and VRAM (with this error (random read))
I do not know what else to do!
thanks for the page with the codes now precisely control how many beeps.
I do not know how to thank you,helping me a lot


EDIT:36 beeps then shutdown this means ram problem?
I tested the ram I do not mark problems,
I have a problem with the VRAM maybe.

this is the screen of vram test: http://s21.postimg.org/hd6w73vuv/memtestcl.jpg

 

The errors could still be false positives: http://hardforum.com/showthread.php?t=1675582 ,"This app has not been updated since 2010 so all these "errors" are most likely the result of obsolete code incompatible with the new architecture. ". Not to say they are false in your case with your obvious problems, but there is still hope its not the graphics card that's the issue(?).

Did you try switch/take off some of the RAM sticks?

 

^ Yeah this is the next step if you haven't already done so. It may be only one or it may be most of the modules are broken...

 

Your error screenshot is too small. Can't read anything.

 

I was just doing
I changed slots, and ram
now I try and pray

MR.Koala I will try to enlarge it asap

EDIT:
different ram and different slots same result, 36 beeps and shutdown
I can not believe that it is corrupted the VRAM,
yet it would seem to be

 

Would you have, by chance, the possibility of trying another video card on your laptop? It is the only thing missing I'm afraid.

It will be, however, odd if the issue is related to faulty VRAM, as that issue has quite graphical symptoms, such as corrupted textures, distorted polygons and the likes. It can be another component of the GPU going haywire then again.

 

the only way to try another gpu would buy it
I have always the same error 36 beeps from the bios ami
gpu seems perfect, temps, throttle, fps, everything
is crazy, maybe it's the bios or faulty VBIOS

 

To test if it's VRAM related, can you somehow feed some large textures to the GPU to fill up the VRAM, but not push the shaders so the card keeps cool?

 

how do I do this?
Can you explain the steps

 

I'm not knowledgeable when it comes to the Windows environment so I can't get any appropriate testing tools for you.

Any Windows graphics dev here?


Blocking the cooling of the GPU to force a higher temperature without real workload is also a way to test if it's temperature-related. Do keep an eye on the temperature reading though.



Edit: didn't notice #8. Doesn't look like a VRAM issue to me.

@Jaycod:
How did you read the errors from that screenshot? All I can see is a 139*180 thumbnail.

 

understood

 

Can you block cooling to get the temp up to around 85C, and if it doesn't crash, run the RAM/VRAM test again?

 

disconnected the fan of gpu
temp 97 no crash or shutdown, the temperature is rising in a few seconds
memtestcl same result of the previous

 

Read the errors from here http://s21.postimg.org/hd6w73vuv/memtestcl.jpg . It seems to be a window screenshot of a 1080 display (the image itself is of lower resolution, of course).

Running Kombuster/Furmark related tests are causing issues? If so, perhaps the issue is more related to the processing pipeline and not to the VRAM.

A simple test could be an OpenGL program that continuously reads images from disk and binds them as textures. Or another one could be simple adding alot of polygons to the card. However, I have no idea if there are tools specific for this end. Try firing up 3dsmax or maya (or sketchup) and making very complex structures (so much that it slows the system down). If it hasn't blown up by then, try rendering the graph scene to put extra stress.

 

I read the screen perfectly
yes running furmark and and when I'm playing
I'm not able to use computer animation software,
is not there another way?

 

There was a program a few years back called ATI Tool, that was used for diagnostics I believe. Try firing it up. Also, do these issues happen when booting windows in safe mode?

 

If we go the "OS is to blame" route, is there any chance you can boot the machine with a GUN/Linux Live CD/USB, and run a GPU benchmark program (Unigine Mark or something) from there?

The chance of it making a difference is small though.

 

hello
ati tool is a project obsolete and no longer supported,
it also happens in safe mode,

Mr. Koala course I can start a linux distro
I do the test just finished downloading the iso

 

can recommend to me a program to test the gpu on linux?
I use mint 15
uningine heaven is ridiculous and I do not find alternatives to Furmark

 

GPU Test includes FurMark and many other tests.

What made you feel so bad about Uningine?

 

has few options
can not choose the duration of the test
ok download and use gputest

 

under linux I ran gputest
all tests are short
and do not receive problems, due to the short duration
I ran everest on windows
I tested everything except the gpu, no errors
gpu test with everest, pointed me to the usual error
at this point I think the video card, has a physical problem, and is not recoverable,
I've done every test possible, at this point I just hope that by changing the gpu the problem disappears.
I am left with nothing else to do!

 

I agree. Hopefully those problems go away by switching to another video card. Really hope it is not motherboard related :/

 

exactly
I'm terrified that it may not depend on the gpu
I have problems only when I stress the gpu.

 

Keep us posted .

Really curious to know if it is the card. Have you placed your order?

 

not now
in this period I have extra expenses
I also have to choose which gpu to buy
I do not know whether to remain with amd, or switch to nvidia
I honestly do not even know where to buy it
I do not trust ebay, too easy to send products flaky, used or passed off as new.
I am referring to professional sellers like upgrademonkey.
I have a lot to think about

 

Is there any chance you can reach a local computer/parts shop, and pay them a bit for a GPU swap test?

 

ok my new gpu finally is here
I install it when I get back from work hope that my problem disappear
keep your fingers crossed with me
i will keep you informed

same error its not a gpu problem maybe bios problem?

 

no one can help me?
36 AMI bios beeps
scroll lock and num lock start blinding for 36 times,after that shutdown
I tested:
ram
cpu
gpu (brand new)

why does this happen?
please someone help me

 

I've read that you have already updated the bios. I may have missed it, but have you try reinstalling the whole OS?

At this point, it could be a faulty motherboard that is causing this error. This will be your worst case scenario.

This may sound stupid, but since you have two cards now, you can try the crossfire set up. Maybe it will make the problem go away.

 

Hello and thank you very much for your response
I have already tried to change the operating system obviously no effect
at the end after spending money on a new video card, I have spent more to send my PC in for service, I also believe it is the fault of the motherboard, I do not understand what it is damaged only to give me problems with the video card, this point is if I can fix the clevo certainly will make a. sli

 

On the bright side, you don't have to feel bad about buying another video card. Your computer has the ability to run dual video card. When it get it back, set up the crossfire and you will be rocking with a monster. So, just treat it as the cost to upgrade.

As for the cost to fixing the computer, I'm hoping you pay with your credit card. The reason is because some credit card, do provide extended 1 year warranty. If not, then you are out of luck. This is the reason why it is not a bad idea to purchase extended warranty on laptops. It's very fragile, and very expensive to repair. When I got mine, I think it cost me USD$150 to get extended 2 year warranty on my Clevo. Just an FYI, I don't usually buy extended warranty. It's just that computers have a fairly higher rate of failure rate.

 

I paid with paypal
you're right
in fact I would have to extend the warranty
but now it is too late
I take it with philosophy,
Meanwhile, I hope they fix it

 

Yes, and good luck to you. Just another opinion, the extended 2 year warranted makes it a 3 year total warranty. Should the computer failed after the 3 year warranty expire, it's easier to make a decision to dump the laptop and buy a new one. It's a gaming laptop, no one keeps a gaming laptop for gaming after so many years anyway.

Keep us posted.