m17x GPU issues

Hi guys, I apologize if some of these issues have been covered on the board previously, but I have been trying to find a solution to these issue for a couple weeks reading through as many threads as possible, and i am at my wits end.

So 2 weeks ago i took over payments on my brothers M17x-r1. It has a q9000 cpu, geforce gtx260m gpu (currently running nvidia 266.38 drivers), 6GB DDR3 ram

So I first starting experimenting with CPU oc'ing, which i seem to have figured out and seems simple enough. Then i starting experimenting with GPU oc'ing, and then my issues came about. After reading through the boards, I decided that overclocking my cpu to 1280 FSB frequency would be safe, and oc'ing my GPU to 600/1500/950 would also be safe. At this point i decided to try running 3dmark06 to see what kind of scores this unit got. First i ran it at stock speeds and got a score of 8544 which to me seemed low, and even looked low on the 3dmark06 online score chart for similar systems. So then i ran it at my OC'd speeds a couple times, getting scores ranging from 6200-9200, trying different settings with my physx (auto-select thru NST seemed to give the best scores).

So i tried running Dragons age at OC'd speeds on one notch below highest settings. I played for 15 minutes, and my computer crashed. So I waited a bit, and tried again using stock GPU speeds, but left my CPU OC'd. So i played on the same settings for about 10 minutes, at which point i exited the game to check temps, and my gtx 260m was over 100 c. At this point im getting concerned.

So i installed OCCT and ran that test a couple times, first time at OC GPU settings, and after 3 minutes of the GPU test, I was already over 100 c. So i stopped the test and changed my GPU back to stock settings. I ran OCCT at stock, and again in 3 minutes time i was back over 100 c on my gtx 260m.

Now it seems like the fan slightly changes speeds, but it seems like i have heard the fan louder before.

Now I don't know what else to do. Is it an issue with my GPU fan? How can i tell? Is this normal behaviour for this system? Can it not run games like Dragons age at graphics detail: high and texture detail: medium? And if so, would OCCT still push my GPU to 100 c after 3 minutes?

I am still new to all this, and only learned how to do any of this after reading through hundreds of pages on this board. But i am officially lost. It seems like i am not getting optimal performance out of this machine, and if i am, I got hosed. Aleinware tech support is coming to fix my keyboard and motherboard next week, but i dont think those are the issue, motherboard only getting replaced because one USB port is screwed.

I can also post CPU-Z shots if anyone wants. Sorry for the long winded message, just need to relay my whole situation, Thanks in advance for any help

 

Try going into BIOS and deleting all BIOS overclocks and reseting to defaults. OC'ing..especially oc'ing the FSB seems to have unpredictable effects on fan speed. Mine would only run on minimum speed and resulted in higher than normal GPU temps. Also if this lappy has been in use for a while you need to remove and thoroughly clean all three heat sinks with compressed air. Directions how to do this in other threads...but it's not hard and will ensure long life of (especially overclocked) GPUs!

As for bad performance...these machines will eat most games for lunch but good drivers and a stable windows install are critical. If it were me...after reseting all BIOS options to defaults I would:
-clean heat sinks
-turn off PCIE Gen2 in BIOS under ADVANCED tab...GRAPHICS SETTINGS
-do clean windows install from DVD
-reload chipset drivers first
-load NVIDIA driver of your choice....so much to be said here..you have to find a combo that works for you...I have 260m's and use 266.58 but others don't like the verde drivers
-load all other drivers and test
-do overclocks only after getting a stable system at default settings
-read and research here a lot and you will always find someone with the same prob to help you slove it
Welcome to the wonderful world of the hobbyists notebook!

 

Thanks for the reply! I am definitely gonna try this tonight.

As for determining at what my system is most stable, do you suggest OCCT to monitor temps? I have read about determining system stability, but could never tell if this is what everyone meant. And also, should a stable system be able to run the OCCT tests without getting into high temps?

 

So just a quick update.

I cleaned out the heat sink intake, the GPU fan, and all the dust off the card. The Heat sink intake was almost completely covered with dust, knew right away that was the main issue.

Rebooted with stock clock on the CPU and GPU. Ran OCCT, and temps didnt get any higher then 75-80 c after over 10 minutes. Ran 3Dmark06 and got 500 points higher then when i had everything overclocked before.

Funny how its always the simplest solution, and the one your not looking for or thinking about.

thanks again orionz

 

Suhweeeeeet!

System stability can be determined in many ways. A system that blue screens/black screens/freezes during gaming/video playback/benchmarking almost always is pretty much unstable by defininition.

Usually it's the video drivers which are a sore spot especially with R1's and even more especially with 280m's. PCIE Gen 2 seems to be a major contributor to this behavior. See this thread: http://forum.notebookreview.com/ali...idia-verde-drivers-bsods-freezes-m17x-r1.html

Finding the best video driver for your configuration and gaming usage seems to involve a good bit of trial and error. Many like these (slightly modified) drivers : http://forum.notebookreview.com/alienware-m17x/535696-modified-nv-197-52-driver-m17xr1-only.html but they are not Verde and not always great for recent games. I have used the whql 197.16 fairly successfully and am currently on 266.58...the latest from Nvidia. Have fun!