Search Log in FR EN

Programmation GPUDécision (GSI)

FISE GSI Semester 9 Common Core Machine translation

GPU Programming

Teaching unit Decision · GI.S9.ACAD.DE
Course on Celene
5ECTS (EU)
1Coefficient
9:20Hours

Séquencement

C C C TP TP TP TP TP P P P

Learning objectives

Understand the specificities and interest of graphical processor programming. Know how to design and execute calculations on nvidia graphics processor.

Prerequisites

ModuleTeaching unitSemester
Algorithms and Programming 3Basic Sciences 3S3
Algorithms and Programming 1Optional Constitutive Elements 1S1
Algorithmics and ProgrammingBasic Sciences 2S2

Program

Course:

  • CPU architecture and performance improvement limit
  • CPU parallelization
  • GPU parallelization

TP:

  • Implementation of calculations on GPU nvidia with CUDA language
  • Comparison of calculation times with CPU.
  • Comparison of calculation times based on load distributions on GPU cores.

Project: Implementation of an image processing algorithm on GPU and comparison with its equivalent on CPU

Assessment

Project evaluation

Skills

1
Ref.VerbDescriptionLevel
C1_2designKnow how to design and execute calculations on nvidia graphics processor3