# Generic approach to Legacy Fortran code porting on GPU

- Institution: CERN
- Department: CERN Computing Seminar
- Speaker: Dmitry Mikushin
- Year: 2015
- Language: English
- Subjects: Computer science
- Source: [CERN](https://videos.cern.ch/record/3012319)
- 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 this seminar we will present our methodology of unleashing the potential of GPU computing in legacy Fortran codes: to what degree the source code has to be modified to become usable on GPUs, how to turn single-threaded Fortran code into multi-threaded ensemble, how to pack all code into single GPU kernel to minimize synchronization stalls, how to map small loops onto parallel threads with custom directives and LLVM/NVVM, how to retain the preferred Fortran formatted outputs and other questions. The seminar is intended to introduce scientific code developers into techniques beyond the standard CUDA/OpenCL/OpenACC programming methodology, conserve the existing codebase and achieve high GPU utilization.

About the speaker

Dmitry Mikushin holds MS degree in computer science from the Lomonosov Moscow State University. His working experience includes GPUs and CellBE software engineering at NVIDIA (2009-2011), computational science and GPU-enabled compilers at the University of Lugano (2012-present). At Applied Parallel Computing LLC (2011-present) he leads GPU porting and optimization projects as well as delivery of HPC training courses to academia and industry.

.

0:00:00 Slide 1

0:04:03 Slide 2

0:06:01 Slide 3

0:08:29 Slide 4

0:10:32 Slide 5

0:19:15 Slide 6

0:21:51 Slide 7

0:22:21 Slide 8

0:24:13 Slide 9

0:24:23 Slide 10

0:27:39 Slide 11

0:29:25 Slide 12

0:29:29 Slide 13

0:30:37 Slide 14

0:31:03 Slide 15

0:33:45 Slide 16

0:34:18 Slide 17

0:35:46 Slide 18

0:37:32 Slide 19

0:37:59 Slide 20

0:38:03 Slide 21

0:38:24 Slide 22

0:39:25 Slide 23

0:39:36 Slide 24

0:41:09 Slide 25

0:42:03 Slide 26

0:43:40 Slide 27

0:44:51 Slide 28

0:45:54 Slide 29

0:47:28 Slide 30

0:48:31 Slide 31

0:49:35 Slide 32

0:51:06 Slide 33

0:53:28 Slide 34

0:53:50 Slide 35

0:53:59 Slide 36

0:54:12 Slide 37

0:56:09 Slide 38

0:56:42 Slide 39

0:58:19 Slide 40

0:58:49 Slide 41

0:58:54 Slide 42

0:58:59 Slide 43

0:59:03 Slide 44

0:59:39 Slide 45

1:00:39 Slide…

## Lectures

1. [Generic approach to Legacy Fortran code porting on GPU](https://invisiblenetwork.ai/learn/cern/3012319/307fbc14d0b0), 1 h 12 min (CERN's player)