Development of online event selection in CBM

Large track densities together with the presence of a non-homogeneous magnetic field make the online reconstruction and selection of events in the CBM experiment complicated. Track finding in the silicon tracker is based on a cellular automaton. To achieve a high track finding efficiency a Kalman filter fitting algorithm is intensively used within the track finder. After a significant memory optimization and a comprehensive numerical analysis, the Kalman filter based track fitting algorithm has been vectorized using inline operator overloading to be flexible with respect to any CPU family used for data reconstruction. Because of all these changes the SIMDized Kalman filter based track fitting algorithm takes 1 μs per track that is 10000 times faster than the initial version. Porting the algorithm to a Cell Blade computer gives another factor of 10 of the speedup. The cellular automaton track finder of the CBM experiment has been also significantly reworked in order to be SIMDized. The SIMDized Kalman filter based fitting routines have been included into the track finder. The SIMDized cellular automaton track finder shows 1000 times increase of the reconstruction speed with respect to the initial scalar version running on the same Pentium 4 based computer. Such significant speed-up of the track finder has also demonstrated the possibility of on-line data selection at the expected input rate of 107 events per second.