BIT 10.2.1K7 but also the other versions in my opinion... A DUBT

[paolo.moro]

I started testing a car to understand its limitations, then pushed it to the max.
I noticed that if I leave the temperature measurement as CPU-AVERAGE the temperature errors decrease A LOT.
I attach the two files, the first that measures the temperature of ALL cores (with multiple errors it was blocked after 13 minutes) with 8 errors
The second, with AVERAGE measure, gave 3 errors after 20 minutes.
At this point I have a doubt: what should I trust more? On the first test?
But if I look at the core temperature detail (present on both tests) I find that the maximum temperature on the first test involved 4 cores, while on the second it involved 9 (and why only 3 errors)?
I don't know if I managed to explain myself...

[paolo.moro]

I also ask if it is possible to save (in the test) the settings as well.
For example in this case, if I have to take the test again I have to remember what values I had entered in the test.
Or maybe it's already there but I don't see it...
Thank you

[Richard (PassMark)]

When selecting CPU Average, then the Temperature threshold exceeded will only be raise when the average temperature is above the set threshold. If you set it for individual cores, then any core temperature above the threshold will cause the error. In your case, if you selected the Average, then the temperature had to be over 90 when average over 16 cores. It very well may be that only 3 times in your average temperature run that did the average temperature exceeded the set temperature.

Note: There is a Temperature tab in BurnInTest that will show the temperatures of the in real time. You can view the temperature as seen by BurnInTest.

I also ask if it is possible to save (in the test) the settings as well.

The settings are not in the reports (File menu), but if you enabled logging (in the preferences) with Activity Trace level one (1) or above, you could find the some of the settings and duty cycle for the test in the file. The other option is to save the configuration file from the file menu (File->Save Test Configuration or File-Save Test Configuration Text).

[paolo.moro]

Thank you. I understood the temperature management a little better. The temperature window is very important and is one of the ones I look at the most.
But my question is this:
What should I give more importance to? At the average temperature or at that of the individual CORE?
I explain what I do: when a new type of machine arrives, I also use the BIT program to push it to the max and see how it behaves. The tests I do are INTENTIONALLY destructive. So I push processor, ram, disk and graphics card almost to the max and at the same time.
If I configure the program with the average temperature I have no errors (or I have an acceptable number). If I configure with single CORE the errors are many.
Please pay attention: I AM NOT disputing the program, which is very useful for me, but just trying to better understand how to manage it.
Thank you