# Usage of GPU for online data processing – the experience of ALICE and LHCb

- Institution: CERN
- Department: EP Software Seminar
- Taught by: David Rohr, Daniel Hugo Campora Perez
- Year: 2019
- Language: English
- Subjects: Computer science
- Source: [CERN](https://videos.cern.ch/record/3012786)
- Licence: [CERN](http://copyright.web.cern.ch)
- Who can watch: Tweens profiles and older (sources give no ratings; placed by Invisible Network from the source's own words)

In LHC Run 3, the ALICE and LHCb experiments will not rely on hardware triggers. The luminosity will be increased significantly, so that up to 40 Tbit/s will be processed in the online farm.

LHCb runs a full-software trigger and ALICE performs calibration and data compression, for which the full software processing enables unprecedented flexibility and possibilities.

Both experiments foresee a two-stage approach, processing the data during data taking for a first data reduction and storing it to an on-site disk-buffer for a second processing when there is no beam.

In order to cope with the higher data rates, both experiments are exploring alternative hardware technologies instead of traditional CPU compute farms.

We discuss the framework design, implementation details, performance, and results of LHCb's full-software trigger on GPUs "Allen" and of ALICE's online and offline data compression and reconstruction on GPUs.

.

0:00:00 Slide 1

0:04:19 Slide 2

0:04:45 Slide 3

0:05:50 Slide 4

0:06:36 Slide 5

0:06:52 Slide 6

0:07:45 Slide 7

0:08:44 Slide 8

0:10:31 Slide 9

0:11:19 Slide 10

0:12:01 Slide 11

0:12:54 Slide 12

0:14:11 Slide 13

0:14:46 Slide 14

0:17:01 Slide 15

0:17:38 Slide 16

0:18:04 Slide 17

0:18:41 Slide 18

0:19:00 Slide 19

0:20:15 Slide 20

0:20:59 Slide 21

0:21:13 Slide 22

0:22:04 Slide 23

0:22:30 Slide 24

0:23:14 Slide 25

0:23:20 Slide 26

0:23:29 Slide 27

0:23:36 Slide 28

0:24:28 Slide 29

0:24:47 Slide 30

0:24:51 Slide 31

0:25:44 Slide 32

0:26:06 Slide 33

0:26:12 Slide 34

0:26:27 Slide 35

0:26:55 Slide 36

0:28:38 Slide 37

0:29:36 Slide 38

0:29:56 Slide 39

0:30:25 Slide 40

0:30:52 Slide 41

0:31:31 Slide 42

0:32:01 Slide 43

0:32:42 Slide 44

0:33:16 Slide 45

0:34:10 Slide 46

0:35:02 Slide 47

0:35:47 Slide 48

0:36:35 Slide 49

0:37:00 Slide 50

0:37:24 Slide 51

0:37:54 Slide 52

0:38:10 Slide 53

0:38:44 Slide 54

0:39:35 Slide 55

0:40:42 Slide 56

0:41:08 Slide 57

0:41:48 Slide 58

0:43:43 Slide 59…

## Lectures

1. [Usage of GPU for online data processing – the experience of ALICE and LHCb](https://invisiblenetwork.ai/learn/cern/3012786/279c11475cb3), 1 h 11 min (CERN's player)