Name: Fifth INFN International School on: "Architectures, tools and methodologies for developing efficient large scale scientific computing applications" ESC13 - Bertinoro (Forlì-Cesena) Italy 20-26 October 2013
Start: 2013-10-20T16:00:00+02:00
End: 2013-10-26T15:00:00+02:00
Location: Bertinoro

Fifth INFN International School on: "Architectures, tools and methodologies for developing efficient large scale scientific computing applications" ESC13 - Bertinoro (Forlì-Cesena) Italy 20-26 October 2013

from Sunday, 20 October 2013 (16:00) to Saturday, 26 October 2013 (15:00)

Monday, 14 October 2013
Tuesday, 15 October 2013
Wednesday, 16 October 2013
Thursday, 17 October 2013
Friday, 18 October 2013
Saturday, 19 October 2013
Sunday, 20 October 2013

17:00 Registration and Welcome

17:00 - 20:30

20:30 Welcome Dinner
Welcome Dinner
20:30 - 22:00

Monday, 21 October 2013
08:30 Session 1
08:30 - 20:00
Contributions

08:30 Welcome and opening remarks - Mauro Morandin (PD)

09:00 Concepts of performance and efficiency - Peter Elmer (Princeton University)

09:50 Modern processors and related optimisation topics - Part 1 - Sverre Jarp (CERN)

11:00 Modern processors and related optimisation topics - Part 2 - Sverre Jarp (CERN)

11:50 Introduction to Performance tuning tools - Peter Elmer (Princeton University)

14:15 Floating point computation: accuracy, optimization, vectorization (with exercises) - Vincenzo Innocente (CERN)

15:00 Floating point computation: accuracy, optimization, vectorization (with exercises) - Vincenzo Innocente (CERN)

16:00 Floating point computation: accuracy, optimization, vectorization (with exercises) - Vincenzo Innocente (CERN)

16:45 Floating point computation: accuracy, optimization, vectorization (with exercises) - Vincenzo Innocente (CERN)

17:30 Student lightning presentations
20:30 Dinner
Dinner
20:30 - 20:30
Tuesday, 22 October 2013
08:30 Session 2
08:30 - 20:00
Contributions

08:30 Efficient C++ coding (with exercises) - Francesco Giacomini (CNAF)

09:20 Efficient C++ coding (with exercises) - Francesco Giacomini (CNAF)

10:30 Efficient C++ coding (with exercises) - Francesco Giacomini (CNAF)

11:15 Efficient C++ coding (with exercises) - Francesco Giacomini (CNAF)

14:15 The Memory Crisis - Peter Elmer (Princeton University)

15:00 How memory allocation works - Peter Elmer (Princeton University)

16:00 Exercises - Memory Allocations - Peter Elmer (Princeton University)

16:45 Exercises - Memory Allocations - Peter Elmer (Princeton University)

17:30 Student lightning presentations
20:30 Dinner
Dinner
20:30 - 20:30
Wednesday, 23 October 2013
08:30 Session 3
08:30 - 14:00
Contributions

08:30 Exercises (Floating Point, Memory use, C++) - Francesco Giacomini (CNAF) Peter Elmer (Princeton University) Vincenzo Innocente (CERN)

09:20 Exercises (Floating Point, Memory use, C++)

10:30 Exercises (Floating Point, Memory use, C++)

11:20 Exercises (Floating Point, Memory use, C++)
14:00 Session 4: Introduction to parallel computing
14:00 - 18:00
Contributions

14:00 Motivation.... The power wall and the emergence of ubiquitous heterogeneous computing - Tim Mattson (Intel)

14:45 Parallel Computing: basic concepts and vocabulary - Tim Mattson (Intel)

16:00 Parallel hardware: from SMP to GPU to clusters to massively parallel supercomputers - Tim Mattson (Intel)

16:45 Core design patterns of parallel algorithms - Tim Mattson (Intel)
18:00 Social Tour
Social Tour
18:00 - 20:00

20:30 Social dinner
Social dinner
20:30 - 22:30
Thursday, 24 October 2013
08:30 Session 5: Hands on introduction to parallel programming with OpenMP
08:30 - 12:50
Contributions

08:30 Multithreaded programming with OpenMP: The SPMD pattern on the CPU - Tim Mattson (Intel)

09:30 Parallel loops with OpenMP - Tim Mattson (Intel)

11:00 The divide and conquer pattern with OpenMP tasks - Tim Mattson (Intel)
12:50 Lunch
Lunch
12:50 - 14:20
14:20 Session 6: Hands on introduction to GPU programming with compiler directives
14:20 - 19:30
Contributions

14:20 GPU architectures

15:15 Core design patterns for the GPU programmer - Tim Mattson (Intel)

16:45 Programming GPUs with directives: OpenACC and OpenMP 4.0 - Tim Mattson (Intel)

18:30 Evening lecture: Exploiting vector units - Tim Mattson (Intel)
20:30 Dinner
Dinner
20:30 - 20:30
Friday, 25 October 2013
08:30 Session 7: Hands on introduction to GPU programming with CUDA and OpenCL
08:30 - 12:50
Contributions

08:30 The kernel parallelism pattern - Tim Mattson (Intel)

09:30 Basics of kernel programming - Tim Mattson (Intel)

11:00 GPU memory hierarchy and reductions - Tim Mattson (Intel)
12:50 Lunch
Lunch
12:50 - 14:20
14:20 Session 8: Hands on introduction to cluster computing
14:20 - 18:30
Contributions

14:30 MPI and the concept of message passing - Tim Mattson (Intel)

15:30 The SPMD pattern in MPI - Tim Mattson (Intel)

17:00 Programming highly scalable systems: "MPI+X" - Tim Mattson (Intel)
20:30 Dinner
Dinner
20:30 - 20:30
Saturday, 26 October 2013
08:30 Session 9
08:30 - 14:50
Contributions

08:30 Students feedback

09:00 Final examination

11:30 Delivery of certificates of attendance

14:00 Shuttle departure (to Forli' railway station)