Time consumption error studies
Posted: Thu 18 Aug 2022 17:03
Hi Didier,
I am running orbit correction studies and noticed that the time needed to finish is quite long.
Basically, the linac contains 9 RF field maps with several hard-edge quadrupole doublets/triplets in between.
In addition, several steerers and BPMs are positioned along the beam line, amounting to approximately 10 each.
The errors are applied only to the quadrupole X and Y displacements and I'm running for 100 steps.
Each step seems to take around 1 minute to finish, which seems quite a lot, considering the orbit correction algorithm (SVD, right?) is not that exhaustive.
What could be the bottleneck?
Cheers,
Lennert
P.s.: Attached is the .ini file.
I am running orbit correction studies and noticed that the time needed to finish is quite long.
Basically, the linac contains 9 RF field maps with several hard-edge quadrupole doublets/triplets in between.
In addition, several steerers and BPMs are positioned along the beam line, amounting to approximately 10 each.
The errors are applied only to the quadrupole X and Y displacements and I'm running for 100 steps.
Each step seems to take around 1 minute to finish, which seems quite a lot, considering the orbit correction algorithm (SVD, right?) is not that exhaustive.
What could be the bottleneck?
Cheers,
Lennert
P.s.: Attached is the .ini file.