MAX-P and MAX-N

Autonomous Machines Jetson & Embedded Systems Jetson TX2
1

Hi everyone,

For the MAX-N is quite straightforward to realize that the mode is meant for full clock speeds. The MAX-P mode is however not clear to me. Could you please give me a brief insight about the difference?

Thanks big time.

2

I suggest these links, for me was really useful.

https://devblogs.nvidia.com/jetson-tx2-delivers-twice-intelligence-edge/
https://devblogs.nvidia.com/jetpack-doubles-jetson-inference-perf/

Best,
Raffaello

3

Hi Raffaello,

first of all thank you for your answer.

I have read these blogs but they do not give a definite description of MAX-P.

Thank you anyway.

4

Hi maycondouglasd,

The MAX-P is for power and performance balance, depends on your use case to do the evaluation to decide if need to use all cores(A57 + Denver), ARM A57 cores or Denver cores in different freq for specific requirements.

Cheers.

5

Hi kayccc,

In NVIDIA Jetson TX2 Delivers Twice the Intelligence to the Edge | NVIDIA Technical Blog it is stated that:

“Max-P, the other preset platform configuration, enables maximum system performance in less than 15W”.

The max-clocks mode still manages to give better performance than MAX-P under the 15W budget, according to the measurements in the blog post.

This is why the definition of what MAX-P really does is still fuzzy.

I would appreciate a slightly more detailed explanation.