qBraid
Challenge Accepted: Boost Our Code with Cython!
CPython is renowned for its optimization capabilities, offering potential for significant performance enhancements. It's time to tap into that potential! We're inviting you to integrate CPython anywhere in our codebase and demonstrate a measurable improvement in performance.
The Challenge
Integration: Use CPython to optimize the time and/or space complexity in any part of the qBraid-SDK codebase. While improvements are welcome across the board, insiders know that the qbraid.transpiler module might just be the hidden gem you’re looking for 👀 💎 .
Demonstration: Provide statistical evidence of performance enhancements, focusing on execution time or memory usage.
Documentation: Include benchmarks, performance comparisons, and your methodology in your pull request.
What should our priority be: optimizing for time or space? Which trade-off is less costly for Qbraid?
@Hope-Alemayehu Both are great but I would say optimizing for time, if I had to choose!
@Hope-Alemayehu How are things going? If you're finding the transpiler difficult to optimize, the qbraid.interface.circuit_equality module is another greater opportunity for improvement with CPython. Matrix calculations can be extremely costly at large dimensions.
I was working on the transpiler and made some improvements. While I understand how to analyze the time complexity and the nature of the improvements, I found it challenging to provide statistical evidence. Would it be acceptable to present the analysis using Big O notation?
@Hope-Alemayehu Yep that would be great!
In addition, you could use time.perf_counter() to give a ballpark measurement of the actual speed-up in milliseconds for a given operation, comparing the old code to the new. For example:
import time
# Start the timer
start_time = time.perf_counter()
# Code to measure
sum = 0
for i in range(1000000):
sum += i
# End the timer
end_time = time.perf_counter()
# Calculate the elapsed time
elapsed_time = end_time - start_time
print(f"The code took {elapsed_time} seconds to execute.")
Hi @ryanhill1 @Hope-Alemayehu! Can I take this issues up to integrate CPython in the code base?
Hi @ryanhill1, Thanks!
I wanted to ask for a clarification: I believe you mean Cython instead of CPython, as the former provides the c-extension to Python which gives time/memory improvements.
from this discussion:
So CPython does not translate your Python code to C by itself. Instead, it runs an interpreter loop. There is a project that does translate Python-ish code to C, and that is called Cython. Cython adds a few extensions to the Python language, and lets you compile your code to C extensions, code that plugs into the CPython interpreter.
@shubhamkaushal765 Yes, that's what I meant! Thanks for the clarification