- author = "William B. Langdon and Justyna Petke and Aymeric Blot and David Clark",
- title = "{GI} Software with fewer Data Caches Misses",
- howpublished = "ArXiv",
- year = "2023",
- month = "6 " # apr,
- keywords = "genetic algorithms, genetic programming, genetic improvement, SBSE, linear representation, software resilience, automatic code optimisation, tabu, nonstationary noise, perf, world wide location, plus codes, zip code, OpenCV, image segmentation",
-
URL = "
https://arxiv.org/abs/2304.03235",
-
video_url = "https://youtu.be/uitJFYFb2P8",
- size = "10 pages",
-
abstract = "By their very name caches are often overlooked and yet
play a vital role in the performance of modern and
indeed future hardware. Using MAGPIE (Machine Automated
General Performance Improvement via Evolution of
software) we show genetic improvement GI can reduce the
cache load of existing computer programs. Operating on
lines of C and C++ source code using local search,
Magpie can generate new functionally equivalent
variants which generate fewer L1 data cache misses.
Cache miss reduction is tested on two industrial open
source programs (Google Open Location Code OLC and Uber
Hexagonal Hierarchical Spatial Index H3) and two 2D
photograph image processing tasks, counting pixels and
OpenCV's SEEDS segmentation algorithm.
Magpie patches functionally generalise. In one case they reduce data misses on the highest performance L1 cache dramatically by 47 percent.",
-
notes = "long form of \cite{langdon:2023:GECCO} Also known as
\cite{langdon2023gi}
Slides: http://www.cs.ucl.ac.uk/staff/W.Langdon/gecco2023/langdon_gecco2023_slides.pdf Poster: http://www.cs.ucl.ac.uk/staff/W.Langdon/gecco2023/langdon_gecco2023_poster.pdf",
