E6400 overheating throttling

I hardly doubt that a conflict is possible. Programs runs on their own virtual machines, so that you don't have any conflicts from any shape or form. No special Supervisor API calls is being performed, like an anti-virus requires. You just call a normal API call to get the data.

Both was done. I'll post my finding soon.

 

The high performance timers within the CPU that ThrottleStop uses are not a protected resource. Any software can access and reset these timers, any time they want to. Software that does that can interfere with other monitoring software trying to get some accurate data from these timers. That's why I recommend that you should not over monitor your computer.

Everest and HWMonitor used to use these timers as if they were the only programs using them and would randomly reset them. To the best of my knowledge, HWMonitor was fixed so it no longer causes problems for other programs. I have not tested Everest recently to see if it has been fixed.

CPU-Z also uses one of these timers and last time I checked, it was using it as a shared resource so it was not causing any problems for other monitoring software.

 

So HWMonitor and TS is safe?

 

I just downloaded and tested HWMonitor 1.16 and it is using the system timers in a friendly manner so it won't cause any problems for ThrottleStop.

That part is OK but by default, it is using the wrong TJMax for my T8100. It is using TJMax=100 when the Intel documentation shows that it should be using TJMax=105.



Intel® Core?2 Duo Processor T8100

This setting will cause the core temperature to be reported 5C lower than it actually is. You will have to go into the hwmonitorw.ini configuration file and set that correctly for your CPU if you want HWMonitor to report your CPU core temperature accurately. A P8400 should also be using TJMax=105

Edit: RealTemp isn't a bad monitoring program.

http://www.fileden.com/files/2008/3/3/1794507/RealTempBeta.zip



It correctly reports Super Low Frequency Mode (SLFM), it correctly shows when Intel Dynamic Acceleration mode is being used, it can report VID voltage, by default, it uses the correct TJMax for the mobile CPUs, it can monitor Nvidia GPU temperatures and keeps track of the minimum and maximum, it can report in the Thermal Status area if the CPU has triggered any thermal throttling since you powered up and it will also show if thermal throttling is in progress and last but not least, it's one of the few programs that monitors the Clock Modulation setting of the CPU. As we've seen, this is very important to detect the throttling schemes that Dell uses.

Almost forgot. You can also set it to trigger an alarm based on CPU or GPU temperature or have a computer shut down based on any alarm temperature you choose. It can test your temperature sensors to see if they get stuck and it can output log data to an Excel file for further analysis. It has an accurate and repeatable single threaded benchmark, it has a mini mode available and you can also adjust the color scheme to whatever you like.

Another Edit: The programmer of Core Temp has decided to start reporting Clock Modulation too. Head here to get the latest version.

Core Temp 0.99.7.
http://www.alcpu.com/CoreTemp/

 

I have followed this thread (and the original Dell Forum thread) since discovering last year that my Latitude was throttling. First, a sincere thank you to unclewebb, Goodbytes, HerrKaputt, John Ratsey, and all the others devoting time and energy in exploring this problem and its solutions.

To add another story, however incomplete:
- using a 65W power supply leads to throttling (accoring to Resource Monitor) so I now use only the 90W or higher power supplies.
- when working out in central Senegal (temps > 40C) in March (bios A20), May (A20) and June (A25) the computer became unusable because of throttling. CPU speed goes to 7% according to resource monitor.
- blowing the dust out of the cooling fan outlet helps but even after cleaning throttling occurs when its hot outside
- my colleagues using other brands of laptops (lenovo, Acer, HP ...) were able to continue working despite the heat. no survey of what CPUs their machines had.
- improving air circulation (laptop cooling fan/pad for the field, table fan plus cooling pad when docked) has helped but not when in the heat.
- throttlestop now installed but have not been back to the field.

I have the laptop for work (analysis, database devel, document prep, etc.) and have minimal time to trouble shoot ... especially if work is disrupted when the computer becomes unusable! I will contact Dell when back in the States to see what they can do. Again, I am very grateful for the time you have put in to try to overcome these problems ... I have used Dell's for over a decade and found the Latitude to be a solid machine protected by Complete Care ... though alas that does change and other Dells I have tried were not so robust (XPS M1330).
Thanks again.

 

This is normal, as the laptop require more power than 65W to operate. To be able to at least run, everything goes under minimal power so that a 65W PSU can be used, but if your system would go to full speed, it would shut-down on you do a lack of power. Also with a 65W PSU, you have to choose between using the laptop, or recharging your battery at normal speed.

I would argue that 40C ambient is too hot for a laptop to operate, and that Dell was probably aiming for long lasting machine rather than pushing the processors to max in term of heat and if it breaks, well too bad for you.
My guess, is that did this because the laptop comes with 3 year warranty minimum, so they try to avoid replacing broken CPU's and motherboard which is the most expensive part of the laptop.
So I guess 1 year warranty devices is beneficial, as most will not get an extended warranty and the laptop will break soon after (under 3 years), and kinda force you to buy a new machine.

 

so what do you guys think. what is that is controlling throttling, is it the bios or windows?
has anyone noticed throttling under linux?

they told me to send them the laptop, the problem is that i am in europe now. arghhhhhhhhhh

 

I'm 99.9% sure that the throttling is managed at the BIOS level. It's caused by rules written by Dell, not Microsoft.

But you should be able to arrange for service at a Dell European service centre. Business people expect a global warranty service.

John

 

that's what i said, but they told me my contract is not next day on site so doesn't include intl warranty. I will try to call Dell Italy.

or what should I do

 

Sorry Unclewebb, I've been swamped with work.

My laptop has been suffering from serious overheating, the GPU has been hitting 105ºC. Dell is going to swap the thermal pad (the guy said thermal paste but I'm pretty sure it's the same old thermal pad) and the heatsink.

I think it is in our own best interest to do tests only on machines that are 100% okay according to Dell. There's no point in saying "hey, my laptop is throttling" when it is overheating too. So I'll run the tests this weekend, after the machine is no longer overheating.

I've been reading the thread sort of, but just to be 100% sure, here's what I plan to test:

- run two CPU benchmarks (probably wPrime since it's what you guys used)
- run a GPU benchmark at the same time (Furmark, for the same reasons)
- run throttlestop and log stuff

Is the above correct? If yes, which stuff should I log with throttlestop?