Mar-13-2025, 08:34 AM
Hello!
I am currently doing my computer science thesis on Python energy consumption. This entails measuring the energy consumption of a Raspberry Pi during the execution of a Python benchmark, and then comparing results across Python versions, implementations, OS, chip architecture etc.
I want to make sure that the benchmark is representative of real world applications for the sake of result validity. I am currently planning on using the "apps" benchmark package of the pyperformance tool, but I am worried it might be too lightweight to represent larger modern applications.
My questions then, are as follows:
If you are a Python dev or contributor, I would love your thoughts on this.
If you are neither on those things, please feel free to contribute any knowledge you have regardless.
Thanks in advance!
I am currently doing my computer science thesis on Python energy consumption. This entails measuring the energy consumption of a Raspberry Pi during the execution of a Python benchmark, and then comparing results across Python versions, implementations, OS, chip architecture etc.
I want to make sure that the benchmark is representative of real world applications for the sake of result validity. I am currently planning on using the "apps" benchmark package of the pyperformance tool, but I am worried it might be too lightweight to represent larger modern applications.
My questions then, are as follows:
- Do you think the apps benchmark of pyperformance are sufficiently representative of real life applications?
- If yes, please elaborate.
- If no, are you aware of any application that would be more suitable, and lends itself to be run as a contained benchmark?
If you are a Python dev or contributor, I would love your thoughts on this.
If you are neither on those things, please feel free to contribute any knowledge you have regardless.
Thanks in advance!