{"guid":"4542455a-5147-2020-2020-202020202020","title":"Analysing colliding particles with 256 Threads - Performance of an embarrassingly parallel problem on the Intel Xeon Phi Knights Landing","subtitle":null,"slug":"EBEZQG","link":"https://c3voc.de","description":"Ich habe ein Jahr lang am CERN meine Masterarbeit zum Thema Vektorisierung/Parallelisierung eines CERN-Photon-Pattern-Recognition-Algorithmus geschrieben. Das Hauptaugenmerk lag auf Optimierung der Performance durch Vektorisierung/Parallelisierung, sodass der Algorithmus auf einem Intel Xeon Phi Knights Landing (64 Kerne à 4 Hyperthreads) möglichst performant läuft.\n\nKurz gesagt, ich habe versucht ein \"inherently embarrassingly parallel problem\" mit OpenMP, Vector libraries, Variablen im Speicher an Cacheline-Grenzen ausrichten, Mathe-Tricks, etc so schnell wie möglich zu kriegen.\n\nDieser Vortrag gibt eine Einführung in die Welt der manycore-Systeme aus Sicht eines Informatikers.","original_language":"deu","persons":["chrysh"],"tags":["mrmcd17","107"],"view_count":169,"promoted":false,"date":"2017-09-01T00:00:00.000+02:00","release_date":"2017-09-02T02:00:00.000+02:00","updated_at":"2026-01-15T17:45:11.112+01:00","length":1956,"duration":1956,"thumb_url":"https://static.media.ccc.de/media/conferences/mrmcd/mrmcd17/107-hd.jpg","poster_url":"https://static.media.ccc.de/media/conferences/mrmcd/mrmcd17/107-hd_preview.jpg","timeline_url":"https://static.media.ccc.de/media/conferences/mrmcd/mrmcd17/4542455a-5147-2020-2020-202020202020-timeline.jpg","thumbnails_url":"https://static.media.ccc.de/media/conferences/mrmcd/mrmcd17/4542455a-5147-2020-2020-202020202020-thumbnails.vtt","frontend_link":"https://media.ccc.de/v/EBEZQG","url":"https://api.media.ccc.de/public/events/4542455a-5147-2020-2020-202020202020","conference_title":"MRMCD 2017 - Bundesdatenschau","conference_url":"https://api.media.ccc.de/public/conferences/mrmcd17","related":[{"event_id":4531,"event_guid":"4a584e33-5551-2020-2020-202020202020","weight":12},{"event_id":4532,"event_guid":"375a5150-574b-2020-2020-202020202020","weight":5},{"event_id":4533,"event_guid":"4a4b5748-4645-2020-2020-202020202020","weight":6},{"event_id":4539,"event_guid":"384d5955-3745-2020-2020-202020202020","weight":9},{"event_id":4540,"event_guid":"55585442-5342-2020-2020-202020202020","weight":7},{"event_id":4542,"event_guid":"56484c54-534e-2020-2020-202020202020","weight":8},{"event_id":4544,"event_guid":"45465a39-3747-2020-2020-202020202020","weight":6}],"recordings":[{"size":148,"length":1956,"mime_type":"video/mp4","language":"deu","filename":"mrmcd17-107-deu-Analysing_colliding_particles_with_256_Threads_-_Performance_of_an_embarrassingly_parallel_problem_on_the_Intel_Xeon_Phi_Knights_Landing_hd.mp4","state":"new","folder":"h264-hd","high_quality":true,"width":1920,"height":1080,"updated_at":"2017-09-02T11:35:05.135+02:00","recording_url":"https://cdn.media.ccc.de/events/mrmcd/mrmcd17/h264-hd/mrmcd17-107-deu-Analysing_colliding_particles_with_256_Threads_-_Performance_of_an_embarrassingly_parallel_problem_on_the_Intel_Xeon_Phi_Knights_Landing_hd.mp4","url":"https://api.media.ccc.de/public/recordings/19307","event_url":"https://api.media.ccc.de/public/events/4542455a-5147-2020-2020-202020202020","conference_url":"https://api.media.ccc.de/public/conferences/mrmcd17"},{"size":70,"length":1956,"mime_type":"video/webm","language":"deu","filename":"mrmcd17-107-deu-Analysing_colliding_particles_with_256_Threads_-_Performance_of_an_embarrassingly_parallel_problem_on_the_Intel_Xeon_Phi_Knights_Landing_webm-sd.webm","state":"new","folder":"webm-sd","high_quality":false,"width":720,"height":576,"updated_at":"2017-09-02T12:01:47.429+02:00","recording_url":"https://cdn.media.ccc.de/events/mrmcd/mrmcd17/webm-sd/mrmcd17-107-deu-Analysing_colliding_particles_with_256_Threads_-_Performance_of_an_embarrassingly_parallel_problem_on_the_Intel_Xeon_Phi_Knights_Landing_webm-sd.webm","url":"https://api.media.ccc.de/public/recordings/19308","event_url":"https://api.media.ccc.de/public/events/4542455a-5147-2020-2020-202020202020","conference_url":"https://api.media.ccc.de/public/conferences/mrmcd17"},{"size":59,"length":1956,"mime_type":"video/mp4","language":"deu","filename":"mrmcd17-107-deu-Analysing_colliding_particles_with_256_Threads_-_Performance_of_an_embarrassingly_parallel_problem_on_the_Intel_Xeon_Phi_Knights_Landing_sd.mp4","state":"new","folder":"h264-sd","high_quality":false,"width":720,"height":576,"updated_at":"2017-09-02T12:02:04.342+02:00","recording_url":"https://cdn.media.ccc.de/events/mrmcd/mrmcd17/h264-sd/mrmcd17-107-deu-Analysing_colliding_particles_with_256_Threads_-_Performance_of_an_embarrassingly_parallel_problem_on_the_Intel_Xeon_Phi_Knights_Landing_sd.mp4","url":"https://api.media.ccc.de/public/recordings/19309","event_url":"https://api.media.ccc.de/public/events/4542455a-5147-2020-2020-202020202020","conference_url":"https://api.media.ccc.de/public/conferences/mrmcd17"},{"size":24,"length":1947,"mime_type":"audio/opus","language":"deu","filename":"mrmcd17-107-deu-Analysing_colliding_particles_with_256_Threads_-_Performance_of_an_embarrassingly_parallel_problem_on_the_Intel_Xeon_Phi_Knights_Landing.opus","state":"new","folder":"opus","high_quality":false,"width":0,"height":0,"updated_at":"2017-09-02T12:02:36.613+02:00","recording_url":"https://cdn.media.ccc.de/events/mrmcd/mrmcd17/opus/mrmcd17-107-deu-Analysing_colliding_particles_with_256_Threads_-_Performance_of_an_embarrassingly_parallel_problem_on_the_Intel_Xeon_Phi_Knights_Landing.opus","url":"https://api.media.ccc.de/public/recordings/19312","event_url":"https://api.media.ccc.de/public/events/4542455a-5147-2020-2020-202020202020","conference_url":"https://api.media.ccc.de/public/conferences/mrmcd17"},{"size":29,"length":1947,"mime_type":"audio/mpeg","language":"deu","filename":"mrmcd17-107-deu-Analysing_colliding_particles_with_256_Threads_-_Performance_of_an_embarrassingly_parallel_problem_on_the_Intel_Xeon_Phi_Knights_Landing.mp3","state":"new","folder":"mp3","high_quality":false,"width":0,"height":0,"updated_at":"2017-09-02T12:02:37.709+02:00","recording_url":"https://cdn.media.ccc.de/events/mrmcd/mrmcd17/mp3/mrmcd17-107-deu-Analysing_colliding_particles_with_256_Threads_-_Performance_of_an_embarrassingly_parallel_problem_on_the_Intel_Xeon_Phi_Knights_Landing.mp3","url":"https://api.media.ccc.de/public/recordings/19313","event_url":"https://api.media.ccc.de/public/events/4542455a-5147-2020-2020-202020202020","conference_url":"https://api.media.ccc.de/public/conferences/mrmcd17"},{"size":184,"length":1956,"mime_type":"video/webm","language":"deu","filename":"mrmcd17-107-deu-Analysing_colliding_particles_with_256_Threads_-_Performance_of_an_embarrassingly_parallel_problem_on_the_Intel_Xeon_Phi_Knights_Landing_webm-hd.webm","state":"new","folder":"webm-hd","high_quality":true,"width":1920,"height":1080,"updated_at":"2017-09-02T12:24:40.142+02:00","recording_url":"https://cdn.media.ccc.de/events/mrmcd/mrmcd17/webm-hd/mrmcd17-107-deu-Analysing_colliding_particles_with_256_Threads_-_Performance_of_an_embarrassingly_parallel_problem_on_the_Intel_Xeon_Phi_Knights_Landing_webm-hd.webm","url":"https://api.media.ccc.de/public/recordings/19316","event_url":"https://api.media.ccc.de/public/events/4542455a-5147-2020-2020-202020202020","conference_url":"https://api.media.ccc.de/public/conferences/mrmcd17"}]}