m9750 Owner looking for opinions/tips

IF YOU HATE READING LONG POSTS, HERE'S THE DEAL:

1. m9750 shuts off at random while on World of Warcraft

2. I brought the temps down from 84c down to 70-74c but it still shuts off after playing a while.

Here's the whole story...

I've read a ton of the info here, and its great to see so many other users.

So here's my deal. My m9750 (XP sp2) has been running decent since I got it. At first it was a powerhouse. I was getting some random shut downs at first, and had no clue why. I use the machine mainly for work (photoshop, VS 2005, video editing, etc.) and never played a game on it till recently.

I've had some issues with the computer, and just wanted to start fresh, but AW said they don't take computers for software work, and since I didn't get a respawn CD with the comp, I'm pretty much on my own with it. I want to do this soon, so any tips on a good setup for my use of the machine??

No my current problem really starts with the fact my good friend just started working for Blizzard... and that's where World of Warcraft enters the picture. I clocked in somehwere around 8 days of playing time (as in 8 full days worth of hours on the game). No problems at all. I installed the basic game, then the upgrade to Burning Crusade, then to the Lich King the other day. No probs.

So I'm playing for a couple hours on Thursday, and out of nowhere my machine just goes OFF. Just the LED for the plug icon is still blue? Its not shutting down, not making a noise, just poof, off. Lots of you know about this, as I've read a bunch of your posts here.

I get the machine back on, get WoW going, then a couple minutes later... OFF. So I read on these forums about what's happening, and download CoreTemp. I'm running 82-83c on both cores. I figure that seems hot, so I turn on a little office fan and point it at the fans. I see at best the temps are down to 79-80c. I downloaded the driver recommended by nVidia and installed the nVidia Control Panel Version 1.5.2400.24. Installed the driver (ForceWare Version 175.32) and changed the power settings on the comp for best performance in the "Always On" setting. Found the controls for SLi and enabled it.

I read about bad heat sink paste, and being the guy I am, I tore the machine apart, pulled the battery, pulled both heat sinks and fans, cleaned the crap out of the board (just with air) and waited for some Arctic Silver 5 to be brought over. Next morning, I cleaned the old crap off, and did a nice thin layer of AS5, bolted everything back in, made sure it was perfect, and turned my machine back on.

At idle I was running 65c and after a couple hours of WoW, I was running 70-74c MAX with no fan pointed at the comp this time, no cooling pad, nada. I played for a few hours, and no probs. Then... OFF. This time when I tried to reboot it would give me a couple blinks of the LEDs, then OFF.

Now what? I'm on the machine right now, but I know with WoW it'll shut off after a while, be it hours or minutes. I also noticed there was two white stripes with a pulsing green bar and the letters SLi, all of which was about 50% transparent to the left of the screen ONLY when I was on WoW, not when I alt tabbed out.

Sorry for the book here guys, I just wanted to lay it all out nice and clear. HELP!

 

It could be your mobo but honestly I have no idea. How about your CPU temps?

 

What should I download to find out?

 

You think there is a better driver to try? I'm downloading 169.28 right now.

 

uhm... try with SLI off?

the m9750 has fairly powerful GPUs as far as I can recall, so using only one, you shouldn't have too many problems maintaining the same detail settings, or anything... might slow down in high-demand areas though.

just turn it off, no other changes should be necessary.

I say this, because if ONE of the GFX cards is bunk for whatever reason, disabling SLI should fix it... if not, you could try swapping the main and secondary cards, and trying a single gpu setup again on what used to be the secondary card... see if that fixes it.

might also be power-related... does the battery still work? does the A/C connection show wear? does the AC ever drop randomly?

 

You can't swap the GPU's. One is a PARENT card and the other is a DAUGHTER. Disabling SLi could tell him if it is his SLi or the Daughter card.

The program for all temps is HWmonitor

 

Thanks guys! The only problem I see with messing with SLi, is that when this problem started thursday morning after TONS of hours of play on WoW, I had never once enabled SLi yet.

I only turned on SLi for the first time thinking maybe it'd balance the load better or something? I'll download HWmonitor right now. Do you think I should revert back to 169.28? drivers?

 

My readings from the HWmonitor are:

ACPI: 74c
Duo T7200 72c (both)
GPU 63c
GPU 70c
HDD 38c

 

erm, you shouldnt need to disable SLI, you bought it with SLI so you've got to use SLI , thats why you shelled the money for!

try the newest driver and tell us how it turns out, because the latest one reduced my temps while maintaining same performance

EDIT: thats while playing or while idle?

 

Sorry for not reading the whole post, got a problem reading long ones, cant concentrate after a while so tell me if i missed anything crucial

Was wondering if those temps are at idle or under full load. Do you feel air coming out of the laptop ? (from the fans)
And did you ever cleaned it to get rid of the dust ?

Sounds like an overheating problem to me, probably CPU's temp going up too high and shuts itself. So that might be a "few" reasons for that.

 

Yeah I addressed all that in the first post, its not a heating/dust prob. I just installed the new chipset, and it looks like my temps are up a bit. That reading was at Idle, just a couple websites open.

Is there any reason that new chipset would give me issues? I have WoW running now just to see what happens. I have garage and car work to do, so if it shuts off, I'll give it a couple hours to relax before trying to turn it back on just to browse here.

 

Well from what i can see it is a heating problem. Running the CPU at that temps at idle means the temps are skyrocketing when you play games. Hence why it shuts down.

Idle temps should be around 35-50 degrees depending on specs ofcourse.

 

What Temp is the CPU reading? I posted the GPU.

 

Those temps, 72c at idle is too high, well same goes for the GPU's too but those can handle a bit of heat in contrast with your CPU which in some laptops are forced to shut down at high temps.

 

If those are your idle temps, your CPU is way too high.
Mine idles at 54C.

When you cleaned the the fans out with compressed air, did you spray between the copper fins? Theres a ****e load of dust there if you've had your m9750 for a while.

Also that Artic Silver 5 is swag. With that 200 hour curing process. Get yourself the slightly better Artic Cooling MX2 that works out of the tube... its cheaper as well.

Also did you remove the northbridge thermal tape?

If you require more cooling, could get yourself a Zalman cooling pad for a couple more degrees off.

 

I agree he shouldn't NEED to, but to isolate the problem, it's a good idea... y'know. debugging?

you know what that is right?

 

shut off after my last post, lol.

 

I'm testing with undervolting right now. Anything else to check to get my CPU temps down? I've seen them steady at around 70c, which you guys say is pretty high? Why would it be so high?

 

Dust.
Fans might not work properly.
Might need new thermal paste.

But since both of you CPU and GPU have high temps at idle id say check the first two.

 

I'm Dust free. I just laid down some AS5. Dropped about 10c from the CPU. Still shut off while on WoW after about an hour or so playing.

 

Just to make sure that its a temp problem try keep HWmonitor running while playing wow and after 10 mins or so minimize to check the temp.

If its still as high like before (which it should since the laptop is shutting down) then you can try replacing the thermal paste (you are going to find if the fans are working properly too while doing it) or send it back to the manufacturer if you still got warranty.

 

...or just use Rivatuner and watch the overlay in-game

 

Mine idles at around 65 in a warm area (undervolted). 70 without undervolting. One point I did have the computer beep at me after playing an intensive game for 2 hours that it was overheating. Nothing after that but occasionally the video driver would crash on that one game and cause a further system crash. And this is very important: this issue only happened in XP.

This is my current setup that causes me no problems:
1. Fresh install of Windows Server 2008 64 bit (free download and 240 day free trial)
2. Newest Video drivers (Nvidia 180.48, or pick your choice)
3. Newest Audio Drivers (Realtek 2.10)
4. Do NOT install the Intel chipset drivers.

Just to clarify what happened to me. My laptop dual booted with both Alienware's installation of XP and my own installation of Server2008. The XP had problems from the beginning. Random memory corruptions, errors while booting, failing to get to the login screen even after the loading screen finished, etc. The list goes on, but that XP just didnt like my laptop. So I finally removed it and the Server OSs are always rock hard stable.

Further notes:
1. The video driver DOES make a difference in how hot the cards run. You have to read the boards to find out which ones are running cards hotter. There is no general rule to follow.

2. Running at max resolution makes the cards run hotter. Even lowering the resolution from 1920 to a 16XX resolution could solve my heat problems.

3. Running in SLI makes the cards run hotter. Especially for games where SLI does not help. So if you are having problems and you know SLI is not worth it TURN IT OFF!

And a note to those replacing the thermal pads with Artic Silver 5. It had been posted on the board here that some heatsinks need the thermal pads (they will have an indent in them). So dont replace everything with Artic 5.

Happy Gaming!

 

Thanks so much. I kept my thermal pads, and just swapped the hard thick stuff with AS5. I'll keep SLi off then. Hey what settings do you have your multiplyers at? If you have the m9750 that is

 

Hey Fly! Glad I could help and hope it really does solve your problem!

As for my CPU multipliers, 6.0x is as low as I can go (makes 1ghz) and truth be told for normal work I keep it locked this low. Nothing I do, not even running VM Ware virtual machines makes me need any more power than that. I jump it to 10x when decoding a Divx film, and I run full blast for gaming. I also have a notebook cooler, as the air ventilation under the m9750 isnt that high off the table so i force some more air underneath it.

 

I'm running cooler now with the multipliers set back a bit. But I still know that I should be able to run a friggin game on this machine with issue? I gotta get AW on the case.

 

I tell ya the truth: move to Vista x64 or Server x64. With today's drivers it is a much more stable platform. And with Server 2008 being free for 240 days you really should give it a try. Alienware does some really weird tweaks to their XP installation and it always caused problems for me, and it really might solve your problem. Of course try to get Alienware to do something, but if they dont help you always have this option.

Sorry if this seems a little daunting. Im a OS tweeker guy and I love installing new OSs

 

Just a little info WOW is not an SLI game and unless they changed something to address that it, it will disable SLI during strart up.

 

WoW still doesnt support multiple GPU's but you can still see an fps increase when running it on an SLI/Crossfire system since it doesnt really disables one card.

 

all this talk had me check my temps which i haven't done in a while.

i installed those new chipset drivers as well, and though i haven't noticed any problems whatsoever, my CPU and GPU temps went up a little bit though. the CPU went up more than the GPU.

my CPU used to idle at around 47ºC. it is now idling at 60ºC.

my GPU used to idle at around 58ºC it is now idling at 65ºC.

i've never replaced the thermal paste, and i've had the notebook for about a year now. i use a NC-1000 for the cooling pad.

would new thermal paste really drop my temps back down into the 40's? now i'm a little curious how my system is running while gaming.

 

Did you clean it from the dust the year that passed ? I would start from there

 

haha yes. i do that every 1-2 months.

 

Whats your temps under heavy load after 10 mins ?

 

after heavy load for 10 minutes:

I ran the HL2 lost coast stress test a few times for the load.


CPU - 71º C

GPU - 77º C

i guess it isn't too bad. though i did notice my frames dropped a lot. i used to get around 130 FPS and now it rendered around 97 FPS on everything high, 1920 x 1200, 2XAA and 4xAF. that isn't terrible, but eh, maybe it's the new drivers too.

 

only 10c difference ? thats pretty good tbh. Dont think you need to do anything else just yet

 

FlyRide - I would call Alienware and ask for a replacement set of Heatsinks both for your CPU/GPU and your other GPU. Even if your out of warranty, the cost of getting them should be around 50 bucks.

I've replaced my GPU heatsink and it brought my temperatures down by like 10C under load, 5 C idle. I'm just waiting on my GPU/CPU heatsink assembly to get replaced now.

Part numbers are the following:
[Item:AREA-51 m9750 (W835DI1) CPU/VGA HEAT SINK, Qty:1]
[Item:AREA-51 m9750 (W835DI1) SPS-ASSY NB VGA HEAT SINK AB042, Qty:1]

I used arctic silver the first time as well, but getting new replacements heatsinks did a better job even with using the Alienware's standard paste.

 

Im confused Heatsinks arent metal things attached to the cpu/gpu which "absorb" the heat ? Hows replacing them lowers the temps ? Instead of just cleaning them that is (and replacing the paste since you are on it too )

 

do they have higher quality heatsinks?

 

While I do agree with your statement, I can say that cleaning it out with air and reapplying arctic silver appeared to do nothing for me, however, actually changing the assemblies (heatsink and fans) did. Maybe fan wear and tear, maybe the heatpads weren't making contact anymore.. maybe I just Eff'ed up my arctic silver application several times in a row. I don't know. All I can say is that when I found that I started to overheat and couldn't do anything about it, I replaced the heatsink assembly, it worked for me.

 

Started thinking about downvolting to be honest. It has no disadvantages and can lower your temps from 5-25c... which is a lot...

 

Levenly, with your temps, I would say your fine at the moment because that what I'm getting right now. Before, however, I was idling at 70C and getting over 100C when gaming.

After replacing just the VGA heatsink assembly, my load temps dropped to 88C max and idle 65C. Running lost coast like you did and my temp was only 76C max. The one I haven't replaced yet still idles at 70C and got as high as 88C in lost coast though. I expect that to change as soon as I get my replacement.

Again, I don't know if it's a better heatsink or not. Just that when all else failed, I replaced it and it worked.

 

CoolMine - How does one undervolt the CPU? what program can be used to do that? Right now, I use Notebook Hardware control to throttle my CPU down to 1Ghz and also keep my temps cooler overall when not gaming but that's about it.

 

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

 

Thanks man!

I just started with a voltage 1.125V (from 1.225v) and I've already shaved off 5C off my load temp(From a max of 82C to 77C). Still testing...

EDIT.

Okay, I didn't shave off too much more. I was able to reduce my max voltage to 1.112V which in turn lowered my max CPU temp to 75C. So a 7C reduction in temp overall for 'free'. I like it.

 

I ended up talking with AW for a while. I had to download some program that writes the ISO CD so that I could load up a new BIOS file. Seems to have fixed the WoW issues that were shutting the comp down. I also reverted back to the old chipset, and reinstalled the video drivers. I actually wiped the video drivers, and the screen went black. After much playing in Safe Mode (which took forever to even get to) it seems to be doing good while gaming, yet I've had at least 4 freezes today. Even the mouse froze, and I had to hold down the power button to shut it off, but never while gaming.

Still don't like my temps. I think I'm gonna reinstall XP, reload the chipset and drivers again, and maybe only load half the programs I have on this machine now. I've also had some malware screw me up, and reg edit programs prolly left some damage while fixing that issu.

 

Are you using the original video drivers? The mouse not moving sounds like an old video driver issue people used to have around here.

 

i have the beta DOX 180.84 drivers...

my gpu temperatures were maxed at 91ºC and 86ºC, and they were in the mid 80ºC while playing L4D.

should that peak at 91ºC concern me?

 

At the moment, no, I don't think you need to worry about it, however, once it starts to hit 95ish, then clean out the fans, heatsinks and possibly reapply arctic silver. I'm guessing you're hitting 91C on the primary card or Core 1 (not core 0)?

I don't know why, but Dox's 180.84 drivers feel very sluggish to me. I guess everyone is measuring performance in Farcry and Crysis only. For me, Mass effect, Assassin's Creed, Call of Duty 4 all took a performance hit using his driver.

 

i'm only playing L4D at the moment, and sometimes fallout 3.

i've pretty much maintained the same frame rate in L4D, though, it's mostly more stable.

i'm guessing it's the primary card, as it's the top one listed. on HW monitor, my cards are listed separately and they don't specify which is which.