hammer
commented
1 year ago Some potential citations:
- Numba: a LLVM-based Python JIT compiler (2015): the Numba paper
- Lessons learned from comparing C-CUDA and Python-Numba for GPU-Computing (2020): experience report of Numba vs C on the GPU.
- Comparing Python, Go, and C++ on the N-Queens Problem (2020): preprint that compares Numba to Go and C++ for a compute-intensive task.
- NPBench: a benchmarking suite for high-performance NumPy (2021): benchmarks Numba against other Python accelerators, maybe not quite what we want.
- Sarkas: A fast pure-python molecular dynamics suite for plasma physics (2022): this paper does a detailed comparison of Numba versus the same algorithm in C.
- Powering Practical Performance: Accelerated Numerical Computing in Pure Python (2022) is another potential option, though I can't get the full text to confirm.
jeromekelleher
Very interesting piece here on comparing Python against C++ and fortran implementations of classic physics simulations: https://www.nature.com/articles/s41550-021-01342-y
We should certainly mention this in the introduction/discussion. "There is still a tacit assumption for many that compiled code in languages such as C and C++ is necessarily more efficient than implementations using interpreted languages such as Python. However, [discuss this paper]."