As a first step towards understanding and after optimizing the 2D cluster finder (CF) is to compare it to the existing CF.

To have an apple to apple comparison I rad the reconstruction with both CFs on the HLT machines that ensures the same CPU for all the processes and both CFs were ran on the same DAQ files.
DAQ files correspond to 2016 run days 125 and 126.

2D stands for the new cluster finder algorithm.
1D stands for the current cluster finder algorithm.

O2 stands for the O2 oprimized compilation of the algorithm.

The results is shown below:

Two cluster finders prodice different number of hits:

In order to check if both cluster finders perform the same way for the lower altro cut (that we had before 2015) same time profiling was done on AuAu2014 data. Here are the restuts :

The above result once again shows that both cluster finders time performance is very similar.

I also looked at the track reconstruction efficiency for the AuAu embedding sample for both Sti and StiCA track reconstruction algorithms. The result is shown below :

It shows that new CF gives slightly higher reconstruction efficiency that the current one.

The ratio of StiCA 2d and StiCA 1d efficiencies is shown below: