Your cluster already records the energy of every job.
Researchers see wall-clock time and core-hours, but many schedulers also record energy per job, and almost nobody looks.
The UK's ARCHER2 supercomputer lets users look up the energy of every job. GreenCode is exploring checks like this for jobs on shared clusters.
GreenCode helps tackle this by measuring research code's energy before and after a change, pointing to the code behind each hotspot, and checking that the optimised code still gives the same results. More on this.
Your cluster may already record the energy of every job. Researchers watch run time and core hours. But many job schedulers also record the energy each job used. And almost nobody looks. You cannot cut what you never look at. Here is the fix you can use today. First, check if your scheduler records job energy. On Slurm, energy accounting records it, and the sacct command shows it. Then compare energy per result between versions, using tools like nf-co2footprint for Nextflow. For a big campaign, estimate before you run, using the Green Algorithms calculator. The UK's ARCHER2 supercomputer lets users look up the energy of every job. GreenCode is exploring checks like this for jobs on shared clusters. GreenCode helps tackle this by measuring research code's energy before and after a change. It points to the code behind each hotspot, and checks the answers stay the same. That was GreenCode Tech Tip number fifty-six. Follow for the next tip, and find more on greencode.ai/tech-tips.
Share this
