New M6500 Discussion Thread

So I guess basically what you are telling us here is that Dell will deliver the M6500 with 8GB modules in the future. It's a great news that those modules will be available soon, but I don't see why you treat this as an achievement made by Dell.
This is not meant to be offensive and I know of your contribution to this community, but I'd find it more honest if you would just state that bigger modules will be available and that it's not Dells credit but the logic consequence of overall technology progress (which Dell in turn will make available to their customers).

Sorry, but since Dells false advertisement on the M6500 with Red Hat I'm kind of allergic to Dells propaganda.

 

I just had my system kill itself and next time I turned it on it told me it had been hit by the thermal trip -- I was playing a fairly regular game (Flatout 2), and the laptop is standing on a hardwood table with free access to the fans...


Any idea what the thermal trip temperature is? I ran speedfan the moment it booted up, but it was hovering at around 60 degrees -- I think I'll try to keep an eye on it..
Anyone else hit by the thermal trip? Seems a little fast to get that when the laptop has been used for less than two weeks :|

 

Assuming that your fans work properly I don't think that this "thermal trip" was triggered by the CPU. I have the same CPU like you and even with 29°C ambient temperature and full load on all cores (including HT) I never exceeded some mid 60°C of CPU temperature. The mobile Nehalems are designed for emergency shutdown at 107°C.
On the other hand the thermal link of the graphics card seems to be a bit underpowered to me. Or at least the BIOS is a bit dull to trigger the fans based on GPU temperature. I had some OpenGL failures (but no total shutdowns or hang-ups) when I let the BIOS control the fans and played some games. If I control them manually there is no problem.
I have the FX 2800M, so I guess in your case this effect should be even more considerable.

 

I just turned on my computer, and will now leave it on this glass table (completely flat, so should provide excellent ventilation) for the next 15 minutes doing nothing.

I am unsure which part of the computer triggered the thermal trip -- it just states in the BIOS log that the thermal trip was triggered (meaning the sensors detected too high temperature on *something*). I think it was the graphics card, though. When you say manual control... which application do you use to control fans? I used a tool for my old M65 laptop once, but that is not supported for the M6500, and I think I'd like to try it out.


Whoa! I didn't finish this message, but can already see the temperature on the computer being a little high.. just Google Chrome open, nothing else that really drags power (CPU usage hovering at around 2%-3%), and this is the temperature as of now:

[/IMG]
(yes, SpeedFan shows a different CPU usage, and I am unsure of whether that is because I was tabbing around or what, for it idled at 2%-3% before)

Does anyone else think that's a little high for doing nothing much?
I'll leave it for a while in a moment, and next time I come back I'll take a new screenshot of the TOTAL idle temperature (I will kill everything I can first)

 

Right, this is what I got after leaving it with nothing running -- I pressed the ALT+PrintScr combo first (which takes only a screenshot of current window) but decided I wanted to include the Task Manager, så I took a second PrintScrn -- thus the 11% CPU usage. It was idling at 2% for ten minutes previous to my taking this screenshot..



So what do you think? Too hot? *sigh* Don't want to call Technical Support again..

 

I use i8kfan from the i8kutils to control my fans.

Depending on your ambient temperature (You're from Denmark, right? So I assume that you have something near 30°C.) I wouldn't consider 55°C idling as too hot.
The values with 60°C+ are a bit raised for an idle system, but far from being alarming. They seem to be caused by some background activity.

If you want to really stress your CPU start 8 calculators, enter big numbers (e.g. 1mio.) and let them calculate the faculty (n!). Just ignore the warning messages spit out by the calculators and let it run for 10-15 minutes. If you don't get another thermal trip you can be sure that it's not your CPU. I guess that you'll not even see 80°C.

 

I am going to install Linux within the next few days -- I have not been home much since I got this laptop, so I have not had the chance to install anything but the stock x65 Windows 7 -- then I'll have a bash
at i8kutils.

As for the temperature, it is around 23 degrees celcius inside here, and I think idling at 50-60 is a little hot to be frank.. Well, I'll find a stress test program, since it does not seem to wish to do your calculator idea -- it simply states invalid input, though I am likely doing it wrong

[ edit! ] Running a Prime95 stress test, where the CPU is maxing out at 100% -- will leave it for 30 or so minutes and report back with temperatures [ / edit ! ]

 

I currently have 26°C ambient and 52°C CPU while the CPU is occupied by a Virtual Machine with 100% spread over all cores (if you consider 800% to be the maximum on 8 cores).
A reason for my lower temperature might be, that my main board was replaced right after I got my M6500. This required the technician to renew the heat paste of my CPU. Maybe he just did a better job on this than Dells factory.

I guess I'm just not up to date. The calculator trick worked fine under XP and I guess MS just made the tool a bit cleverer by now. Maybe decreasing the number would help. But Prime95 will have the same effect.

Edit:
I hope you don't mind that I borrowed some of your signature elements to create my own one.

 

I meant full usage, 100% on all cores -- or 800%/100% if you wish to be pedantic

I suppose you might have gotten a better bit of pasting, yeah -- I wonder if it'd be worth the hassle of opening it up and putting new paste on.. They are madly simple to work with these machines (though the M6500 is a whole lot more bothersome than, say, the M65).

I don't think it's a massive issue, though, for I could never get the temperature over 69°C when running Prime95 maxing out all cores. I think, however, that I need to do a test of the GPU instead. That's where my issue lies, I believe.

 

I just put all cores to 100% for 15 minutes and had constant 68°C in the last 10 minutes.

I don't think this is reasonable. Since our highest temperatures are virtually identical, I think my pasting isn't better than yours.
I'm still running BIOS A00. Maybe later versions have a different fan control behavior which leads to higher idle temperatures. Or the 3800M might produce more heat than the 2800M, even when idle and heats the CPU via the heat pipes.
My graphics card currently has 54°C and is idle.