JetPack 4.2.1 Low Power Consumption ?

Autonomous Machines Jetson & Embedded Systems Jetson Nano
1

Dear all,

I did some test on the new JetPack 4.2.1 where Nvidia said it has lower consumption than the last version.
This is the results I had (Accurate Power Lab Supply set at +5Vcc/3A):

BOOT SEQUENCE:
Old Version → 450mA
New Version → 1050 mA

STAND BY MODE:
Old Version → 280mA
New Version → 350 mA

OPENNING BROWSER:
Old Version → 495mA
New Version → 1150 mA

So, I was surprised to see those metrics !
Because my project will works on batteries, I think I will go back to the old JetPack Version.

Do you have any feedback on this point ?
Because Nvidia said : Lower consumption…Really ?

Thank you for your replies.

Best regards,

2

Hi christophe.herault,

Would you please help to let me know where the statement of “Lower consumption” you saw?
This could help to clarify and find the right information for Jetson Nano power consumption improvement.

Thanks

3

Hi,

I think he might be talking about the JetPack 4.2.1 release notes, section 1.2 :
https://docs.nvidia.com/jetson/jetpack/release-notes/index.html

Better handling of insufficient power supplies
[b]Boot process updated to require less power
[/b]Widget to display under-voltage warning
4

Hi Baramuse,

Yes you are right !
I didn’t understand why Nvidia said that the boot process require less power…It is the opposite !
But that’s ok, I will do with that.

Best regards,

5

Hi Christophe,

Here is power measurement result (average current) of Jetson Nano devkit for JetPack 4.2 and JetPack 4.2.1.
From the results it shows there are no power differences on STAND BY MODE, and energy consumption is reduced during bootup.

BOOT SEQUENCE:
4.2 → 531.2 mA, 43 sec to bootup to Ubuntu desktop => 115.09 Joule energy consumption
4.2.1 → 589.3 mA, 32 sec (-11 sec) to bootup to Ubuntu desktop => 94.28 Joule (-20.81 Joule)

STAND BY MODE: (1 minute average)
4.2 → 260.5 mA
4.2.1 → 260 mA (-0.5 mA)

Please note that the above result are measured with Ethernet disabled and Ethernet cable plugged out, to ensure getting stable power data.
The reason is Ethernet activities differs time to time, and it will causes unstable power variance.

Could you elaborate more details about operation steps of test methods, how you take power measurement, and which peripherals connected to Jetson Nano devkit?
Thanks.

6

Hi Christophe,

Could you elaborate more details about operation steps of test methods, how you take power measurement, and which peripherals connected to Jetson Nano devkit?

This could help to clarify if there is any abnormal on devkit or SW issue.

Thanks