Introduction to Parallel Programming with CUDA: "How to plow a field with 1024 chickens?", Belgrade, Serbia

Monday 18 Feb 2013, 10:00 → 14:45 Europe/Belgrade

Library (Institute of Physics Belgrade)

Training Evaluation http://survey.ipb.ac.rs/index.php?sid=65543

10:00 → 10:05 Opening

Speaker: Dr Antun Balaz (Institute of Physics Belgrade)

10:05 → 10:10 Overview of the training

10:10 → 10:55 Introduction to heterogeneous parallel programming

Speaker: Petar Jovanovic (Institute of Physics Belgrade)

Slides Introduction_to_heterogeneous_parallel_programming.pdf

10:55 → 11:40 CUDA C overview

Speaker: Dusan Stankovic (Institute of Physics, Belgrade)

Slides CUDA_C_overview.pdf

11:40 → 12:00 Coffee Break

12:00 → 13:00 Hands-on session: Tiled matrix multiplication (using shared memory)

Speaker: Petar Jovanovic (Institute of Physics Belgrade)

Slides cuda-mm-example.tar.gz Tiled_matrix_multiplication.pdf

13:00 → 14:00 Hands-on session: Reduction on GPU

Speaker: Dusan Stankovic (Institute of Physics, Belgrade)

Slides Reduction_and_numerical_libraries_CUDA_codes.tar

14:00 → 14:10 Coffee break

14:10 → 14:40 Hands-on session: Using CUDA numerical libraries

Speaker: Petar Jovanovic (Institute of Physics Belgrade)

Slides Using_CUDA_numerical_libraries.pdf

14:40 → 14:45 End of training