Phase-2 mkFit: propagation to plane / Kalman operations on plane / Matriplex with support for scalar operations and VDT (superseding #148) #151
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
mmasciov
force-pushed
the
prop-plane-rb1-plus
branch
2 times, most recently
from
f9e6c06 to
b48568d
Compare
mmasciov
force-pushed
the
prop-plane-rb1-plus
branch
2 times, most recently
from
be6c619 to
4c28443
Compare
mmasciov
changed the title
|
I'm not able to reproduce the second set of warnings in RecoTracker/MkFitCore/src/KalmanUtilsMPlex.cc -- it might be a false positive. I was running: I'll review things very carefully once more ... and I probably need a newer release-base. |
….cc), move structs BeamSpot and DeadRegion out of Hit.h.
…ify the new propagateToPlane code.
* Add VDT support to Matriplex, mostly to work on Matriplex scalar types.
- Functions are prefixed as fast_xyzz(), same as in VDT.
- Controlled by define MPLEX_VDT. Additionally, if MPLEX_VDT_USE_STD is also
defined, the fast_xyzz() functions fall back to using std:: variants.
This is useful for performance comparisons.
- Add reduction operator and an assigner class / method to extract scalars
(one i,j element) or to assign to it.
* Massage propagateToPlane low level implementation
- Use the new Matriplex functionality to simplify code.
- Remove the nmin, nmax indices initially introdcued to support GPU code.
…ements. Define TBB_DISABLE in Makefile.config to make it so.
- PropagationMPlex.cc - prop-to-R - PropagationMPlexEndcap.cc - prop-to-Z - PropagationMPlexPlane.cc - prop-to-plane - PropagationMPlexCommon.cc - applyMaterialEffects, common multiplication functions. Move code from PropagationMPlex.icc into corresponding files above. Replace templates that were used for CUDA testing a long time back with explicit types.
…inal transcription into Matriplex operators.
…S flags, keeping option to modify them for R&D and/or testing purposes
mmasciov
force-pushed
the
prop-plane-rb1-plus
branch
from
bfab41c to
62bccea
Compare
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
PR description:
As per title, this PR supersedes PR #148.
This PR is focusing on mkFit developments for phase-2.
While originally meant to be purely technical (mkFit is not yet enabled by default in phase-2), a small bug was fixed that is slightly affecting the phase-1/Run-3 performance. This was reported at Tracking POG on October 7, 2024. Changes for Run-3 are anyway negligible.
For phase-2 mkFit, this PR:
Propagation to plane is used as a mkFit-default choice for phase-2 (where mkFit is not deployed yet), being required to select hits on tilted layers, while phase-1 configuration is left unchanged (no significant gain wrt. current default).
In addition, standalone mkFit functionalities are also extended/improved.
PR validation:
Please, refer to presentation at Tracking POG on October 7, 2024 (including MTV results).
--> For phase-1 (Run-3), (only physics) performance is only slightly affected, due to a small bug-fix in application of material effects (commit 4a6088d).
--> For phase-2, mkFit physics performance (mkFit is not used by default so far) is largely improved.