# I/O for Deep Learning at Scale

- Institution: CERN
- Department: other events or meetings
- Speaker: Quincey Koziol
- Year: 2019
- Language: English
- Source: [CERN](https://videos.cern.ch/record/3002563)
- Licence: [CERN](http://copyright.web.cern.ch)
- Who can watch: Teens profiles and older (sources give no ratings; placed by Invisible Network from the source's own words)

Deep Learning is revolutionizing the fields of computer vision, speech recognition and control systems. In recent years, a number of scientific domains (climate, high-energy physics, nuclear physics, astronomy, cosmology, etc) have explored applications of Deep Learning to tackle a range of data analytics problems. As one attempts to scale Deep Learning to analyze massive scientific datasets on HPC systems, data management becomes a key bottleneck. This talk will explore leading scientific use cases of Deep Learning in climate, cosmology, and high-energy physics on NERSC and OLCF platforms; enumerate I/O challenges and speculate about potential solutions..

0:00:00 Slide 1

0:00:07 Slide 2

0:01:12 Slide 3

0:02:30 Slide 4

0:04:45 Slide 5

0:04:49 Slide 6

0:05:00 Slide 7

0:05:36 Slide 8

0:07:42 Slide 9

0:09:00 Slide 10

0:10:36 Slide 11

0:10:50 Slide 12

0:11:29 Slide 13

0:12:19 Slide 14

0:13:00 Slide 15

0:14:09 Slide 16

0:15:25 Slide 17

0:17:23 Slide 18

0:18:55 Slide 19

0:19:27 Slide 20

0:20:14 Slide 21

0:20:36 Slide 22

0:21:35 Slide 23

0:22:49 Slide 24

0:23:08 Slide 25

0:23:39 Slide 26

0:24:02 Slide 27

0:24:37 Slide 28

0:26:14 Slide 29

0:27:09 Slide 30

0:28:26 Slide 31

0:30:06 Slide 32

0:30:19 Slide 33

0:31:36 Slide 34

0:31:44 Slide 35

0:31:54 Slide 36

0:31:59 Slide 37

0:32:27 Slide 38

0:33:20 Slide 39

0:34:11 Slide 40

0:34:28 Slide 41

0:35:23 Slide 42

0:36:07 Slide 43

0:37:21 Slide 44

0:38:43 Slide 45

0:38:48 Slide 46

0:40:13 Slide 47

0:41:52 Slide 48

0:41:56 Slide 49

0:43:19 Slide 50

## Lectures

1. [I/O for Deep Learning at Scale](https://invisiblenetwork.ai/learn/cern/3002563/9c0f45678b17), 44 min (CERN's player)