Dear Devs,
As in the title, I think it would be useful to have processing times for the most demanding stages (not as detailed as Performance Benchmarks | CryoSPARC Guide, which is also a bit outdated) listed in the (future) case study pages for more benchmarking possibilities than the built-in mechanisms and to help with estimation of resources required for real-life size datasets.
Hi @bsobol for comparing our HPC cluster vs. workstations and other cryosparc instances, I found the job “Extensive Validation” it is not perfect, but gives you an idea if something is different between instances, nodes, resources, or even if running per-GPU.
another options is to have a workflow / blue print for a standard data set.
It is good to see most of the advanced jobs run.
Yes, that’s what I meant by “built-in mechanisms”.
Thanks for the suggestion @bsobol. We will consider inclusion of some running time information and hardware information in future case studies. However because your system may be different than ours, you may also want to compare newer benchmarks that you run on your system with benchmark results on systems similar to the systems used in the case studies. That comparison should help you predict case study running times on your system, based on running times on our system(s).
