T3 remove redundant new

[YonsiG]

In the T3 kernels, we have some functions called runTripletDefaultAlgoBBBB, and also the BBEE and EEEE regions. Those cuts are duplicating with the passPointingConstraintBBB ... functions. On the other hand, introducing beta iteration cuts to triplet objects are not physically meaningful. In this PR, we still keep the first beta cut as geometric requirement and move to passPointingConstraint function, but remove all the beta iteration functions.

Another improvement is that the triplet radius is calculated and saved for each of the T3s. In this case, we don't need to calculate the radius of triplet multiple times when building T5, pT3 and pT5s. This will also bring timing improvement. Slides describing the influence of this PR is summarized here. https://indico.cern.ch/event/1407246/contributions/5914814/attachments/2846057/4976175/check%20the%20T3%20cuts%20EEEE%202.pdf

[YonsiG]

closed the previous PR with thread divergence. created a new one. Addressing Slava's first comments already.

[YonsiG]

/run standalone /run CMSSW

[github-actions[bot]]

The PR was built and ran successfully with CMSSW. Here are some plots.

OOTB All Tracks

The full set of validation and comparison plots can be found here.

[YonsiG]

Merging this PR will be able to close Issue #214 and Issue #247 related to triplet building cuts

[YonsiG]

/run standalone

[github-actions[bot]]

The PR was built and ran successfully in standalone mode. Here are some of the comparison plots.

The full set of validation and comparison plots can be found here.

Here is a timing comparison:

   Evt    Hits       MD       LS      T3       T5       pLS       pT5      pT3      TC       Reset    Event     Short             Rate
   avg     43.8    323.0    122.9     69.7     91.5    492.8    128.5    157.5     98.3      3.3    1531.3     994.7+/- 267.6     418.3   explicit_cache[s=4] (master)
   avg     44.2    323.5    121.8     49.1     97.0    496.5    133.2    159.4    100.2      2.7    1527.5     986.9+/- 262.4     416.0   explicit_cache[s=4] (this PR)

[YonsiG]

This is the timing summary running on cgpu-1 after the current change.

This is the timing summary on cgpu-1 of the master.

The timing reduction of T3 kernels is ~20%. Due to moving radius calculation from T5 to T3, the T5 kernel time is also reduced by ~12%.

[slava77]

This is the timing summary running on cgpu-1 after the current change.

please rerun to check that s=8 is just a fluctuation

[YonsiG]

This is the timing summary running on cgpu-1 after the current change.

please rerun to check that s=8 is just a fluctuation

Yes, I updated the previous comment by replacing the timing info

[slava77]

/run standalone /run cmssw

[github-actions[bot]]

The PR was built and ran successfully in standalone mode. Here are some of the comparison plots.

The full set of validation and comparison plots can be found here.

Here is a timing comparison:

   Evt    Hits       MD       LS      T3       T5       pLS       pT5      pT3      TC       Reset    Event     Short             Rate
   avg     43.1    323.0    124.8     71.1     93.3    544.9    129.6    159.2    105.1      3.1    1597.3    1009.3+/- 266.4     436.8   explicit_cache[s=4] (master)
   avg     45.6    323.4    123.7     50.6     98.5    548.8    135.9    158.9    108.9      2.4    1596.6    1002.2+/- 264.6     433.4   explicit_cache[s=4] (this PR)

[github-actions[bot]]

The PR was built and ran successfully with CMSSW. Here are some plots.

OOTB All Tracks

The full set of validation and comparison plots can be found here.

[slava77]

/run standalone /run CMSSW

[github-actions[bot]]

The PR was built and ran successfully in standalone mode. Here are some of the comparison plots.

The full set of validation and comparison plots can be found here.

Here is a timing comparison:

   Evt    Hits       MD       LS      T3       T5       pLS       pT5      pT3      TC       Reset    Event     Short             Rate
   avg     43.5    323.7    122.9     70.6     93.7    545.5    131.4    157.0    104.4      3.0    1595.7    1006.7+/- 268.6     435.2   explicit_cache[s=4] (master)
   avg     45.0    323.6    122.1     48.5     97.4    545.1    134.4    157.4    105.7      2.6    1581.9     991.8+/- 267.2     432.5   explicit_cache[s=4] (this PR)

[github-actions[bot]]

The PR was built and ran successfully with CMSSW. Here are some plots.

OOTB All Tracks

The full set of validation and comparison plots can be found here.

[GNiendorf]

Saw this warning while compiling, is this okay? Warning is maybe-uninitialized, not sure if it was discussed already.

[YonsiG]

Saw this warning while compiling, is this okay? Warning is maybe-uninitialized, not sure if it was discussed already.

Thank you for spotting this! I did a check and adds a printout just before line 203, and it can print meaningful values for the betaIn variable. And to be saved to the triplet collection, it has to pass through all the requirements, so for a good triplet the betaIn should be calculated and saved. To be safe, I can add initialization to betaIn as 0, as an extra protection for this.