Parallelizing pipeline runs¶
By default, nipoppy bidsify, nipoppy process, and nipoppy extract will run every participant-session pair sequentially. This can be very slow for large datasets and suboptimal if more computational resources are available. The following guides show how pipeline runs can be parallelized for different computer/server setups.
Automatic job submission on some HPC systems.
Tips for local parallelization on systems with no job schedulers.