Nvidia chip problem

There is no way of telling if your GPU is effected. Until it actually fails. Not good I know. It is distictly possible that you have had the thermal paste poorly applied after your motherboard was replaced, especially as you are seeing such high temps.
Have you changed your drivers since your motherboard was replaced? Newer drivers could, and often do, cause the GPU to get hotter as they make better use of the hardware acceleration offered, not necessarily more efficient use though.

 

bartjeh01
when did you last clean out your cooling system?
what drivers are you running?
what BIOS?
Are you undervolting your CPU?

high temps do not neccesarily mean something is defective (cept cooling paste/heatpipes). But 100C is too much

 

i have >100C too. i cleaned and refitted the heatpipe using AS5. No change in temp so far... i think the heatpipe is way too small for the 8600m gt. my notebook is just 2 months old, so it can't be dust. what bothers me, are the few fins at the end of the heatpipe. in my opinion way too few. the cpu's has dozens, the gpu itself just has 10 or 12. no wonder about those high temps.

i have to admit, even if my notebook goes to 107C (after hibernating), everything runs fine. no thermal throttling or downclocking or even shutdowns. but how long will my notebook be alive with this temps? i already have a zalman 1000 notebook cooler - gave me just a few degrees less...

 

The cooling system is completely diffrent in the 6625 and 6324 you have.
>100C is too much for the 6625, I have 87 (playing AoC), with new drivers and volted CPU. No stutters or anything, or sleepfixes, and runs like a dream (much better than some of my friends desktops )

 

I agree entirely with you on the size of the effective cooling area. Way too small. I have asked Zepto about this and their reply was to ignore my comments and tell me my notebook needed to be sent in for cleaning. I am not prepared to wait nearly 1 month for the notebook to be returned to Zepto, cleaned, and the sent back to me (15 working day turn around quoted to me by Zepto!!!). I had only just cleaned the insides by removing the keyboard and taking the heatsink assy off. My max temp at the moment is 96C, before I cleaned it I regularly hit 102C, not too bad an improvement, but I feel it needs to be lower.
I was toying with the idea of removing the crap finned section and soldering some from a Dell or some other notebook that has a lot more fins, but after the nVidia announcement I decided against completely screwing my warranty.

 

1. About a week ago (I do it allmost every month)

2. 169.09

3. I had z15N015, but with that I could'nt even play for more then 5 minutes, it got a bit better when i downgraded to z15N004

4. Yes, i am undervolting. atm my CPU is at 0,6Ghz and 0,85V but it is undervolted all the way (goes to 1,5V max)

And no, my driver is 169.09 since the day it was released (tough it is not an official Zepto driver)

 

Im running 175.80 (V x32)
015
Undervolted to 0.8500V in SLFM, 1,0125V at 1,8GHz (max)

2 things Im a bit suspicious about your volts:
-004 doesnt support SLFM, wich means CPU cant go to 0.85V, min is something like 0.925V.
-My stock max voltage is 1,1750V so 1,5V sounds awfully much!
You need to check that one, cause ~0.4V is very much

 

RMClock lets me put the lowest voltage at 0,85V, I dont think there is a way to verify this. (and i was wrong on the 1,5, its 1,125)
Maybe i need higher voltages because mine is a T7700 wich goes to 2,4Ghz

And since I am on XP and zepto has't released newer drivers since the stone-age for my GPU, I stick with the latest stable one, thats 169.09..

 

Just to clarify, 0.85V is the bottom of the LFM voltage range in the Intel datasheet:

Each CPU is preset with voltages within those ranges.My T7300 came with HFM=1.25V, LFM = 1.0V and SLFM = 0.9V.

Some lucky people get CPUs with voltages in the lower part of those ranges.

John

 

In RMClock I can choose next to SLFM (Super Low Frequency) 0.85V, that must mean I am one of the lucky ones with a nice undervoltable CPU and a not so nice overheatable GPU..

 

I think all can put the lower fequency voltages to the lowest, but at voltages at high multipliers (high load) needs more power, and some CPUs work at lower voltages than others. (Iv had mine to go under 1V stable at one point)

 

My 6224W spikes over 100 easily too. I dust out too. It's a heatpipe problem...

 

Yes, but does yours stutter while playing a game?

 

Yesterday i undervolted my cpu, taking 0.975V instead of 1.1375V on full load. DeltaT was amazing (>20C!!). I have the feeling that my GPU runs cooler (a few degrees) on full load. Maybe because the cooler does not have to carry out such an amount of hot air.

I am doing a burn-in test right now, but so far i recommend everyone with a hot laptop to do UV.

 

Your GPU will also be cooler when CPU is cooler, as the cooling system has a lower load to take care of.

 

i did a torture test 1h (Orthos), then i played Cod4 about 30 minutes. all was rocksolid and my gpu temp just had gone up to 97C, without undervolting my gpu temp was >105C. Great! Even without sleepfix, i did not notice any stuttering anymore. have to play more to assure that.

 

Temperatures around 100 C is not really great. 50~70 ish is great.

 

5070 is great but thing is GPU temps are so hard to bring down...
Nvidia really screwed up badly...

 

Hi all
I think my 8600 gt just crapped the bed after i left the computer on idle for a bit.
The image just got colourfully scrambled after a while, and the system was non-responsive.
After a few colourfully scrambled reboots I got the machine to work properly again, but now it's back to colourful scrambles.
Did a re-install of xp (working my way through the scrambles) which solved nothing, and my guess is it's a hardware (gpu) issue.
I just read the entire thread, and I gather that it's applicable to my case.
I have a zepto 6224 which I bought August 2007, and I think I reverted to the original drivers and bios at one point.
I'm going to pay a visit to my nearest computer hardware store and order/buy another 8600 gt and just replace the hell out of the old one (if that's at all possible).
If anyone feels it might in fact not be the gpu, or if anyone thinks I should apply for warranty, kindly let me know. I'll be in the kitchen making a sandwich.
Thanks!

 

You cannot replace the GPU in that notebook. You will have to send the notebook back to Zepto to get them to repair it.

 

Will do, thanks!

 

Let us know how you get on with Zepto. It will be interesting to see what they put it down to. Did you get the standard 1 year warranty when you bought the notebook? If so they will probably claim that it is out of warranty, but as this is a GPU that has a potentially known design flaw, they are still responsible. Providing you're in the UK that is, I'm not sure about the rest of the world.

 

The display cable may have gotten loose, if it works sometimes.

Send it to Zepto anyway, at least if its under warranty.

 

I think i did get the 1 year warranty.
I'll keep you posted on their reaction.
This sounds about right:

 

The initial post has been updated

 

Well that is good news

 

That's an interesting statement:

However, when read in the context of this and this, which indicate that the same materials have been used in the GPUs for some time, then the question is how does the Zepto nVidia GPU failure rate compare to the industry average?

John

 

Hi Rene,

I assume you mean the failure rate of models where the GPU is the cause for a needed repair?
The way it's stated now, would include the repairs for all different causes.

Can you confirm that the failure rate is specified to GPU failures only?
Thanks!

 

Well... considering the "industry average" would really be between two companies, i would doubt its validity.

Then again, if it turns out ATI is getting much better success rates...

 

Initial post edited again to be more specific.

 

Why does Nvidia not have some spokesmen or people who could speak on their behalf?

Gamers are what keeps Nvidia and ATI in such good health. Couldn't they at least come clean, once and a while?

 

This problem crept up on nVidia. It's a combination of the way the chips are packaged and the operating conditions. I think it took a long time to figure out why the chips were failing since, it appears, the packaging hasn't changed for some time.

EDIT: The Inquirer has now published more semi-technical explanation of the underlying problem.

John

 

I've just cleaned my fans and even though max temp dont go above 90 , at 84 i'm getting the shuttering. any ideas why?

I was the same before just probably didnt notice it since max went up to 110 sometimes.

Cheers

 

check your CPU temps

 

Update your driver, and check your CPU temps as said above

 

Okey.. regarding everything in this post. Is there a solution for the downclocking stutter problem now? Can I send my laptop which obviously doenst work as intended and get a new, working one?

My 6324w is in total fire and stutter since purchase.

 

flipfire, Sprint - thanks for your replies,

idle temps: CPU 52 GPU 52 celsius
full load(gaming) CPU 97 GPU 94

i do have the latest drivers.

 

CPU 97C???
Oh gosh, you have a problem!

Processor doesnt like to go over 80C, let alone 95C!

Do this tutorial on undervolting
http://forum.notebookreview.com/showthread.php?t=235824

If you dont get your temps down to 80C you need to take a look at the therma paste on the processor. Something is not right here.

edit: My CPU temp is at gaming 70-74C, sometimes lower, depending on ambient and game played

 

thanks sprint, was your CPU temp a lot higher before undervolting? also is 60 Celsius normal for idle? i thought it should be around 30. Ambient is about 22.

I'm only asking because i'm not sure if something is wrong with my laptop or not. Ideally i would prefer a fully working one without me tweaking it to much.

I kind of feel i got a product that wasn't fully functional from day 1, and all this trouble with the shuttering just added to my frustration.

CreX - i don't think there is a solution for the problem at least not a permanent one. Sure you can clean your fan undervolt but the there is a fundamental issue with the GPU, and as far as i know it can't be fixed, it's a design problem.

 

my stuttering was issued by the cpu, not the gpu. i did a check at the temps and my cpu (under full load) was about 87-90 C. in my bios, there's a function to adjust the throttling of the cpu - it was at 87C!. SO i undervolted my cpu, gained nearly 20 C less and my stuttering was gone. i personally never had downclocking issues with my gpu, even on high temps. it was the cpu. now everything works fine. my gpu gets "locked" about 97C without sleepfix, but i did not notice any stuttering or frame jumping.

 

My idle temp is between 55 (where the fan kicks in) and 47. Usually around 54C (Undervolted). Note that I got SLFM disabled. With it enabled I would get lower temps.

Is your fan running all the time when idle? If it does you have a problem with your thermal paste.

If I remember right my CPU temp was around 80 when gaming at normal volts and BIOS 004.

 

....bump....

 

I want to test that as well! How should I do? I'm totally unexperienced in that subject.

 

Crex, you should start by reading Flipfire's ' Undervolting Guide' and/or Ikovac' 'Overclocking and Undervolting guide', both here on NBR.
NBR member Blackbird has also made a guide on undervolting with RMClock with handy screenshots.
Cheers.

 

read trough this

http://forum.notebookreview.com/showthread.php?p=2637648#post2637648

http://forum.notebookreview.com/showthread.php?t=235824

 

Thanks! What is DeltaT?

 

I guess it's delta temperature, that is the difference in temperatur

 

The things " Rene S - Zepto" has written is this thread, what exactly does it mean for us? Can we get a new, working laptop out of warranty or what?

 

If it breaks when in warranty yes ofc. He just stated that the failure rate isnt any worse than on other models (older with nVidia chips)

 

Okey, but what about the overheat issue? Does that count as a broken one within the "failure rate"?