Worried About Overheating

Hey guys, this is my first time posting, but I've been a long-time reader of this forum for all my m11x questions. I've having a serious concern though: I think my laptop is overheating like crazy.

When I first got the laptop it was great and everything ran fine. I checked the temperatures initially and they all seemed to be within normal ranges. However, recently my laptop has been really chugging along while playing games, such as SCII or the Sims 3. When I checked the temperatures just couple minutes ago of playing SCII for 45 minutes, it was at 87 degrees! That's way too hot, right?

I have a m11x r1, OC'd to whatever the bios let me. I'm thinking that the main problem is that the laptop is having trouble with the new monitor I got for my birthday--a 22' Samsung HD LCD. Could hooking it up to this size of a screen have such an adverse affect on my temperatures?

Thanks for any and all help!

 

The cpu is rated to hit 105 before shutting off to prevent damage, so 87 is fine.

If it's hotter now then before, it could be too dusty/hairy/dirty inside. I like to clean my laptops every ~6 months regardless of how they're performing.

 

Bizarre, my R1 never goes higher than 56°C, and I do have these games. Try cleaning your fans and vent, maybe reinsall your bios (I'm using the A04). BTW, 87 is comment for many other machines and is OK for the lap to suport.

 

Try elevating the back end or using a cooler?

 

Are you guys talking about CPU, GPU or HDD temp?

I have never seen higher than 65°C in the GPU, and I do game a lot sometimes (Bioshock, L4D, Crysis, etc.)

About CPU, higher I have seen is 62°C and it was while converting an HD video, took like 1 hour to do it (with my old laptop, I needed to leave the lap on all night to do it)

 

I'm using Coretemp to monitor my processors. I'm not sure how to check my GPU heat...I'm actually not that computer savvy. Right now, after attempting to take apart and clean the insides (very little dust came out, btw..not sure what to look for). Temp for both my cores right now are 53-55. I'll game a little bit using just my laptop screen and see how that goes.

Thanks for all the help so far though!

 

Core2 series are going to run hotter as they're 45nm vs the i-3/5 dual-core series, which are 32nm.

I don't like to see any CPU, whether it be mobile or desktop push past 80C for extended periods. GPU is another story.

 

87C is a lot higher than I recall R1 owners reporting for gaming temps. Yes it's within spec, but to me it sounds like what you're reporting is abnormal for the R1.

 

56C CPU gaming, about 60C idle R1 SU7300

I remember it because I thought it was really funny the CPU in idle ran hotter, but the fan does cycle to about 52C ish before letting it creep back up again

In gaming, the fan is full blast, but limited by the GPU pumping lots of heat out.

 

Haha, are you kidding me? I've never hit over 60 and that was even on my older laptop. What the heck are you doing to make your laptop generate that much heat?

 

I'm a bit paranoid of heat, let me admit that up front. I don't overclock my M11x R2, and at home I NEVER run it without a cooling pad underneath it. If I know I'm going on an extended trip I'll bring the cooling pad with me. That being said, the only time I ever had a problem with my M11x was when I tried overclocking it to 166 Mhz... outside of that it's been running rock solid for a week and a half now as my only computer... running through the night most of the time as I've been downloading my games off of Steam and running disk defragmenting software.

I did recently edit my power savings mode to run extremely cool even when plugged in though. The nVidia card is disabled when plugged in or on battery mode, and the processor can't run more than 75% load Max. Call me paranoid... but this laptop might need to last me at least 2 more years before I can build another nice desktop for everyday use.

 

I played the Sims 3 yesterday for about an hour without the 22-inch monitor and it ran a lot better and cooler. I haven't played SC2 yet so I can't tell, but do you guys think that hooking the m11x to an external monitor could really generate that much strain and heat?

I'm thinking about getting a cooling pad anyways. Any suggestions?

 

Are you playing any of the games that I mentioned? If you are, what settings do you play at? Are you using an external monitor?

 

No but that's a bit insane. I've run Crysis/Crysis Warhead on High for awhile and it's never hit over 50, actually.

 

It won't cause additional heat from the GPU core itself... the worst case is that is uses up more electricity to store more pixel locations in the GPU memory...

Unless if you run VSYNC on all your games, and have extremely low settings on all, then you *might* see a difference in heat generated, but other than that, running it at max will not result in any appreciable difference.

 

After playing SC2 for about an hour, I'm only at 66 degrees. Maybe the little dust I blew out of my system did help after all. I turned down the graphics quite a bit...do you think this would affect the temperature? I still can't imagine what made it get up to 89 a couple days ago.

 

It could also be ambient temperature. I'll admit mine does a bit warm while watching flash or playing games but not to the degree my previous laptop got to.

 

A thought just came to me; do you have your cooling set to passive in the Command Center?

 

Yes my record CPU temperature on SU4100 was around 65 C (it's idling usually at something ridiculous around 35-42 C). That was benchmarking on overclocked CPU (1.7GHz) in my apartment when my air conditioner died and the surrounding temperature was around 33-35 C inside. GPU was something like that as well, never above 66 C. 87 C is not TOO high in general, it wont hurt the CPU.... but it seems unusual for m11x-r1. I would make sure first the vents are clear and the cooler is actually blowing the air out.

 

It's blowing air out, and it idles at around 55.

How can I check? I never even knew you can adjust your cooling settings.

 

Another thing: I haven't really been downloading updates for my laptop since getting it. Do you guys think this would help at all? I heard people mentioning the BIOS update, specifically.

Is there an easy way to keep on top of updating your systems? I usually don't check for updates until something goes wrong, haha.

 

Hit FN + Prt Scrn, then go to AlienFusion, advanced settings for your power plan (I'm using Balanced), and then go down to GPU settings, I think. If not, just explore the tabs until you find how it cools. It could be getting hot because it's set to passive, but I'm not sure.

 

Here is where to find the cooling setting if you are using the Windows Mobility Centre (instead of the Alienware Command Centre - although the settings are the same):

 

Ah, thanks for the insight guys. I just checked and they are both indeed set to active.

 

My CPU hits around 72C when running games like shattered horizon or Red Fraction 2 Guerrilla.

I have my CPU overclocked to 160 and no GPU overclock. I'm not worried whatsoever.

The computer normally runs around 50-55C.

This is all while running throttlestop. I'm using throttlestop to get my CPU temp.

 

Using CPUID HWMonitor v1.16.0 with an M11xR1 SU7300 Over-Clocked BIOS A05 nVidia 260.63

• CPU min:34C Max:49C
• GPU min:39C Max:57C
• HDD min:33C Max:36C

​

This is with the nVidia GPU engaged and moderate GPU OC 520/1248/840.

Played Crysis for two hours and the notebook has been on for 10 hours.
The HWMonitor stats are from the last two hours while I played Crysis.

To be honest I don't think I've ever seen a temperature higher than 60C on my M11xR1.

The funny thing is, it used to run even cooler when I had BIOS A01 but it was prone to having the fan remain in an on state and thus try to deep freeze my M11xR1. This 'locked on' fan state has been fixed in BIOS A05.

I have a Tosh 'netbook' that runs at 75C which it has been doing for over five years now since I put Windows XP on it, without any problems.

If you use HWMonitor to monitor notebook temperature and it's consistently running at 85C (without load) then there may be an issue if it's peaking into the 90C range.

If it's just peaking at 90C and averaging lower temperatures, then it's not ideal, but it shouldn't impact the system.

And you believe HDD and CPU activity is less than 5% when the system is idle. I have Apache and MySQL server running in the background on my M11xR1 and without the nVidia GPU engaged the fan hardly comes on and CPU and GPU sit at 40C and 45C respectively.

I don't have any answers, but I figure if you know the facts on at least one M11xR1 SU7300 it'll give you something to go on and a loose point of reference.

 

Thanks DGV, being able to compare my system to anything helps. I guess I'm not too worried about the heating problem now (been a couple of days with no problems/anything going over 66), so I think the problem may have been with my graphics settings for those games. I've turned of vsync (which I believe someone suggested) and things have been running much smoother.

Although pushing 90 might not damage the system in the long run, it's definitely unsettling to have something so hot to the touch. It felt like I could've cooked an egg on the bottom/vent of my laptop.

 

DrGoodvibes: If you are monitoring your core temperature with HWMonitor then you need to go into the hwmonitorw.ini configuration file and make sure it is using the correct TJMax value for your CPU.

CPU_0_TJMAX=105.0

Here's the documentation from Intel.

Intel® Core?2 Duo Processor SU7300 (3M Cache, 1.30 GHz, 800 MHz FSB) with SPEC Code(s) SLGS6, SLGYV

If this value is not correct then your core temperature will not be reported accurately.

 

Thank you uncle Webb.

I had no idea there was a requirement to update the ini file.

Now have a copy of the Mobile Processor and Intel® Core™2 doc and all it's specifications and reading through it. Knowledge is power. ... if one can understand it.

One thing I note is that the processor should not be stored for any length of time over 85C. Remember this is the storage temp. and not operating temp. but given 85C is mentioned, operating at temps over this may also not be optimal. There may be some variance with the 22mm x 22mm package size.

http://download.intel.com/design/mobile/datashts/32012001.pdf

As noted in 3.9 Maximum ratings and seen in table 5

Just some information to add to the melt.

 

Which is what some of us have been saying the whole time. Those of you who advise those to throw caution to the wind and not worry about temps exceeding ~80C, and let the onboard 'thermal-protection' take care of things are really doing an injustice. If you really believe in it, unplug the fan or block the exhaust of your AW, and run it for a week like that, taking video as proof, and post it back here.

This discussion would be moot over at any Desktop-hardware-oriented forum such as Xtreme, [H]ardForum, you name it. Same rules apply gentleman.

 

Heat dissipates very rapidly within these CPUs. At full load, it can drop 25C from the hottest spot on the core to the center of the heat spreader where Tcase is measured. When your peak core temperature is reporting 100C, the majority of the CPU is at 75C or less. The average temperature of the entire CPU will be under the 85C storage temperature specification.

Here's a good picture of what a hot Core 2 looks like from SemiAccurate



I was curious to see how much thermal abuse a Core 2 could take so I did a test. I overclocked my E8400 by 20% and jacked up the core voltage to 1.35 volts. To create some serious heat I decided to run Prime 95 Small FFTs and just to make things interesting, I disabled the fan. Most people would think that it would be only a matter of seconds before it overheated and crashed.

It ran 100% reliably for 3 hours without any errors while running Prime. I finally got bored and stopped this test but the computer was still running just fine. Could it survive a week like this? I don't know, maybe. I certainly wouldn't bet against it still running a week later.

http://img11.imageshack.us/img11/276/hote8400fw5.png

The core temperature rose up to the thermal throttling point and basically just flat lined there. When it got too hot, it would throttle a little to keep it from going higher. When it cooled down a little, it would go back to running at full speed. Intel's thermal throttling scheme is very advanced and this processor management is happening hundreds of times a second to maximize performance without damaging the CPU. Intel generally sets the official thermal shutdown temperature 25C beyond the thermal throttling point but thermal throttling is usually enough to keep the core temperature well within check.

The thermal throttling point for most of the 45nm Core 2 mobile CPUs is 105C so the shutdown temperature isn't until 130C. Intel wouldn't do crazy stuff like this if they didn't believe in their CPUs. They're not interested in losing money from too many RMA returns.

When you read their documents, the engineers are proud of what their CPUs are capable of. They started using individual core temperature sensors and positioned them at the hottest spots on the core because this allows them to run their CPUs reliably at a higher temperature without having to worry about any problems. As long as your CPU is running reliably and is not thermal throttling, it is operating within the Intel design spec.

On a laptop, something else might fail but it's very unlikely that it will be an Intel CPU that fails.

 

Thank you Uncle webb

Your tests have shown a C2D can sustain some pretty intense punishment.

Given my CPU managed to get up to 50C once or twice, I wont be frying eggs on it any time soon. (CPU_0_TJMAX integrated in HWMonitor)

I've been looking for documentation on nVidia GPUs but finding that a bit harder to find.

I'm for wondering how much punishment a nVidia GPU can sustain.

Peak and sustained temperature thresholds on a GPU seem a bit more elusive.

 

Wow, thanks Uncle Webb! It's definitely reassuring that the m11x carries such high-quality parts--internally, at least, haha. Now if someone could fix the hinges...

 

I propose taking the test a little further. My concerns are not outright failure, but degredation of the CPU due to Intel's quoted warnings posted earlier. I propose finding the maximum overclock of the CPU. Instead of stressing with Prime95, stress with Linx or Intel Burntest. Run the CPU like this for a week without proper cooling. I'm betting your maximum overclock will start to degrade.

 

The 85°C storage temperature specification is not relevant when the CPU is powered up just like the thermal throttling specification is not relevant when the CPU is sitting in storage. Both specs are to insure the long term reliability of the CPU.

All I'm saying is that most people worry too much about their core temperature. If a CPU is constantly thermal throttling like it was in my test, then it is not operating within spec and it could degrade over time. If your CPU is not thermal throttling then its core temperature is running within the Intel specification and should be OK.

I agree that CPUs can degrade over time from excessive voltage or heat. I just don't think this should be a major concern for the majority of users that play a game for a few hours and their peak core temperature gets up to a big number like 80°C or 90°C.

On most 45nm Core 2 mobile CPUs and the Core i7/i5 32nm CPUs, PROCHOT# is not asserted until 105°C so at a core temperature of 80°C or 90°C, it is still well within the Intel recommended operating temperature specification.

 

if the system is randomy shutting down should we be worried about the core temps though?

 

Make sure your accelerometer is disabled. Can you be more specific about this?