Influence of environment temperature in Jetson Nano performance

• Jetson Nano B01
• DeepStream 5.1
• JetPack Version 4.5
• TensorRT Version 7.1.3
• Question type
• Requirement details: Relation Envir. temperature VS Jetson performance

Hi, I had tested a Jetson Nano inside a IP65 encolsure box in 2 differente cenarios:

- Indoor: where the environment temperature is stable at 23/24ºC. The temperature inside the box stabilized at 35ºC and CPU/GPU temperatures at around of 50ºC. The resultant average power consumption was 4.0 Watts.

- Outdoor: where the environment temperature is stable at 20/25ºC. The temperature inside the box changed between 40/44ºC and CPU/GPU temperatures at around of 57ºC. The resultant average power consumption was 4.6 Watts.

In both cenarios, the CPU, GPU usage was the same, around of 50%.

My question is:

Why is Jetson Nano DevKit consumming +0.6 Watts outdoor comparing to indoor cenario? Is this +15% of power consumption justified by the higher enclosure temperature of the air? How can I justify this increase, knowing that in both cenarious the software was the same (something based on deepstream test 2 + python bindings)?

Hi, for thermal relevant info, please refer to the thermal design guide in DLC.

You can use tegrastats command line to show the system usage in real time. Basically if same app, same environment temperature and same thermal solution, the result should be almost same too.

The environment temperature is not the same.

The different temperature will cause different component character showing.