| Title |
Application of GPGPUs and Multicore CPUs in Optimization of Some of the MpdRoot Codes |
| Authors |
- A.I. Fatkina, O.O. Iakushkin, N.O. Tikhonov
Saint Petersburg State University, Saint Petersburg, Russia
|
| Abstract |
We analyzed the ways to optimize MPDRoot algorithms using existing solutions from external libraries. We also examined the libraries designed to work with graphics accelerators and multi-core CPUs, such as cuRAND, cuFFT and OpenCL FFT. The paper describes the ways to expedite a portion of Kalman filter by transferring it to GPUs or multi-core CPUs using the implementation included into the MPDRoot package.
|
| Funding |
This research was partially supported by SPbU (Saint Petersburg State University) grants 9.37.157.2014, 0.37.155.2014 and Russian Foundation for Basic Research grant (project no. 16-07-01113). |
| Paper |
download WEPSB025.PDF [0.894 MB / 3 pages] |
| Conference |
RuPAC2016, St. Petersburg, Russia |
| Series |
Russian Particle Accelerator Conference (25th) |
| Proceedings |
Link to full RuPAC2016 Proccedings |
| Session |
Poster Session B |
| Date |
23-Nov-16 16:50–18:30 |
| Main Classification |
Modern trends in accelerators |
| Keywords |
framework, GPU, distributed, detector, experiment |
| Publisher |
JACoW, Geneva, Switzerland |
| Editors |
Maxim Kuzin (BINP, Novosibirsk, Russia); Volker RW Schaa (GSI, Darmstadt, Germany) |
| ISBN |
978-3-95450-181-6 |
| Published |
February 2017 |
| Copyright |
|
