1 minute rule vs modern computers
Posted: Sun Mar 03, 2024 2:47 pm
After 6 years break, I have returned to the PE to solve one of my missing hard problems. I was motivated by my brother, who recently started to solve problems here.
During those years the environment I am working in has completely changed: 6-9 years ago I was working at rather low-end computer, today I using one of the top ones. My hardware become better, tools and libraries are nowadays better optimized - or better fit my architecture - and as a result my old codes can run even up to 8-10 times faster now. Which makes me feel like cheater. It is much easier to write a brutforce for old problems (which I never did). A program with giant array using 30 GB of RAM? No problem.
My question is: did the PR team take into account the technological progress and when creating new problem do they set greater input numbers to fit today standards of average computational power? What is your hardware setup used to testing and measuring execution time before publishing problem? Your programming language? It would be nice to have anything fixed I can refer to.
During those years the environment I am working in has completely changed: 6-9 years ago I was working at rather low-end computer, today I using one of the top ones. My hardware become better, tools and libraries are nowadays better optimized - or better fit my architecture - and as a result my old codes can run even up to 8-10 times faster now. Which makes me feel like cheater. It is much easier to write a brutforce for old problems (which I never did). A program with giant array using 30 GB of RAM? No problem.
My question is: did the PR team take into account the technological progress and when creating new problem do they set greater input numbers to fit today standards of average computational power? What is your hardware setup used to testing and measuring execution time before publishing problem? Your programming language? It would be nice to have anything fixed I can refer to.