The manual says the shadow core of Tegra Jetson 1 has a lower power, so I want to measure whether this core also leads to a lower temperature. So I took of the fan, and run applications at the core 0 and then run at the shadow core. I don’t know why I always got a higher temperature of the low-power shadow core. Do you have some suggestions? Both cores are full loaded in the measurement. Thanks!
I don’t know the answer, but you may find the specs for the ARM Cortex A15MPCore has information on the topic:
[url]Documentation – Arm Developer
…on that web page open the “Products” in the upper left corner, select “Cortex-A Series”. On the left side you can then pick “Cortex A15 Processor”. Then there is a tab for “Resources”, under which various documents are available…you may need to create a login to download some of these.
Try these :
Jetson/Performance - eLinux.org
[url]http://elinux.org/Jetson/Jetson_TK1_Power[/url]
Look at the section “Restricting to Low-Power Core only”
One reason for the shadow core might be that when you don’t need performance and you are doing just some small maintenance tasks, the shadow core can do them with lower power than the normal cores. If you need to do some heavy calculations (even just on one core), the normal cores might be more efficient.