Frankfurt am Main, Germany
Frankfurt am Main, Germany

Time filter

Source Type

Richter M.,University of Bergen | Aamodt K.,University of Oslo | Alt T.,Frankfurt Institute For Informatik Ifi | Appelshauser H.,Goethe University Frankfurt | And 42 more authors.
Conference Record - 2010 17th IEEE-NPSS Real Time Conference, RT10 | Year: 2010

The ALICE High Level Trigger comprises a large computing cluster, dedicated interfaces and software applications. It allows on-line event reconstruction of the full data stream of the ALICE experiment at up to 25 GByte/s. The commissioning campaign has passed an important phase since the startup of the Large Hadron Collider in Nov 2009. The system has been transferred into continuous operation with focus on the event reconstruction and first simple trigger applications. The paper reports for the first time on the achieved event reconstruction performance in the ALICE central barrel region. © 2010 IEEE.


Gorbunov S.,Frankfurt Institute For Informatik Ifi | Aamodt K.,University of Oslo | Alt T.,Frankfurt Institute For Informatik Ifi | Appelshauser H.,Goethe University Frankfurt | And 39 more authors.
Conference Record - 2010 17th IEEE-NPSS Real Time Conference, RT10 | Year: 2010

The on-line event reconstruction in ALICE is performed by the High Level Trigger, which should process up to 2000 events per second in proton-proton collisions and up to 200 central events per second in heavy-ion collisions, corresponding to an input data stream of 30 GB/s. © 2010 IEEE.


Gorbunov S.,Frankfurt Institute For Informatik Ifi | Rohr D.,Kirchhoff Institute for Physics | Aamodt K.,University of Oslo | Alt T.,Frankfurt Institute For Informatik Ifi | And 48 more authors.
IEEE Transactions on Nuclear Science | Year: 2011

The on-line event reconstruction in ALICE is performed by the High Level Trigger, which should process up to 2000 events per second in proton-proton collisions and up to 300 central events per second in heavy-ion collisions, corresponding to an input data stream of 30 GB/s. In order to fulfill the time requirements, a fast on-line tracker has been developed. The algorithm combines a Cellular Automaton method being used for a fast pattern recognition and the Kalman Filter method for fitting of found trajectories and for the final track selection. The tracker was adapted to run on Graphics Processing Units (GPU) using the NVIDIA Compute Unified Device Architecture (CUDA) framework. The implementation of the algorithm had to be adjusted at many points to allow for an efficient usage of the graphics cards. In particular, achieving a good overall workload for many processor cores, efficient transfer to and from the GPU, as well as optimized utilization of the different memories the GPU offers turned out to be critical. To cope with these problems a dynamic scheduler was introduced, which redistributes the workload among the processor cores. Additionally a pipeline was implemented so that the tracking on the GPU, the initialization and the output processed by the CPU, as well as the DMA transfer can overlap. The GPU tracking algorithm significantly outperforms the CPU version for large events while it entirely maintains its efficiency. © 2006 IEEE.

Loading Frankfurt Institute For Informatik Ifi collaborators
Loading Frankfurt Institute For Informatik Ifi collaborators