News from GPU parameters tuning

So far every measurement taken with --memSize 30 GB and 128 orbit timeframes --> wrong
Retuned the parameters with
- --memSize 15 GB
- 32 orbit timeframes

Dataset	Default sync time	Optimized sync time
pp 2MHz	2658.22 ms ± 4.50 ms	2507.92 ms ± 4.26 ms (5.65%)
PbPb 50kHz	4708.97 ms ± 14.41 ms	4352.53 ms ± 6.35 ms (7.57%)

Question

How I simulate TFs: events = interaction_rate * n_orbits * 0.0000894
For 50kHz 32 orbit I get events = 143
Is it ok?

To-dos:

Tune remaining steps
Test robustness with different datasets with same nature
Tune with other datasets configurations e.g. pp 750kHz
Draw more decent plots
Test on MI100