Name: Second INFN International School on: "Architectures, tools and methodologies for developing efficient large scale scientific computing applications" ESC10 - Bertinoro (Forlì-Cesena) I 22-27 November 2010
Start: 2010-11-21T16:00:00+01:00
End: 2010-11-27T15:00:00+01:00
Location: Bertinoro

Second INFN International School on: "Architectures, tools and methodologies for developing efficient large scale scientific computing applications" ESC10 - Bertinoro (Forlì-Cesena) I 22-27 November 2010

from Sunday, 21 November 2010 (16:00) to Saturday, 27 November 2010 (15:00)

Monday, 15 November 2010
Tuesday, 16 November 2010
Wednesday, 17 November 2010
Thursday, 18 November 2010
Friday, 19 November 2010
Saturday, 20 November 2010
Sunday, 21 November 2010

16:00 Registration and Welcome

16:00 - 20:30

20:30 Welcome Dinner
Welcome Dinner
20:30 - 20:45

Monday, 22 November 2010
08:30 Session 1
08:30 - 20:00
Contributions

08:30 Welcome and Opening of the School - Federico Ruggieri (RM3)

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 Exercises - Introduction to Performance tuning tools - Andrzej Nowak (CERN) Sverre Jarp (CERN)

14:15 Introduction to GPUs/CPU accelerators - Andrzej Nowak (CERN)

15:00 Exercises - Continued focus on Performance tuning tools - Andrzej Nowak (CERN) Sverre Jarp (CERN)

16:00 Exercises - Continued focus on Performance tuning tools - Andrzej Nowak (CERN) Sverre Jarp (CERN)

16:45 Exercises - Continued focus on Performance tuning tools - Andrzej Nowak (CERN) Sverre Jarp (CERN)
20:30 Dinner
Dinner
20:30 - 20:30
Tuesday, 23 November 2010
08:30 Session 2
08:30 - 20:00
Contributions

08:30 Efficient C++ coding - Sebastien Binet (LAL/IN2P3)

09:20 Efficient C++ coding - Sebastien Binet (LAL/IN2P3)

10:30 Exercises - Basic C++ optimisations - Sebastien Binet (LAL/IN2P3) Vincenzo Innocente (CERN)

11:15 Exercises - Basic C++ optimisations - Sebastien Binet (LAL/IN2P3) Vincenzo Innocente (CERN)

14:15 The Memory Crisis - Lassi Tuura (FNAL)

15:00 How memory allocation works - Lassi Tuura (FNAL)

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

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

18:30 Evening Lecture: "The new C++ runtime library: more and better" (Paolo Carlini, Oracle Italia)
20:30 Dinner
Dinner
20:30 - 20:30
Wednesday, 24 November 2010
08:30 Session 3
08:30 - 16:30
Contributions

08:30 How memory allocation works - Lassi Tuura (FNAL)

09:20 Monitoring/Debugging Memory Usage - Peter Elmer (Princeton University)

10:30 Exercises - Monitoring Memory Usage - Lassi Tuura (FNAL) Peter Elmer (Princeton University)

11:20 Exercises - Monitoring Memory Usage - Lassi Tuura (FNAL) Peter Elmer (Princeton University)

14:00 Reducing Memory Usage - Peter Elmer (Princeton University)

14:50 Exercises - Reducing Memory Usage - Lassi Tuura (FNAL) Peter Elmer (Princeton University)

15:40 Exercises - Reducing Memory Usage - Lassi Tuura (FNAL) Peter Elmer (Princeton University)
17:30 Social tour
Social tour
17:30 - 19:45

19:45 Social dinner
Social dinner
19:45 - 19:45
Thursday, 25 November 2010
08:30 Session 4
08:30 - 20:00
Contributions

08:30 Floating point computation - Vincenzo Innocente (CERN)

09:20 Exercises - Optimization floating point computation in C++ - Vincenzo Innocente (CERN)

10:30 Vectorization using SIMD (part I) - Vincenzo Innocente (CERN)

11:15 vectorization using SIMD (part II) - Vincenzo Innocente (CERN)

12:05 Exercises - Vectorization using gcc - Alfio Lazzaro (CERN) Vincenzo Innocente (CERN)

14:30 I/O Efficiency - Gerhard Brandt (DESY)

15:15 Exercises - I/O Efficiency - Gerhard Brandt (DESY)

16:15 Exercises - I/O Efficiency - Gerhard Brandt (DESY)

18:30 Evening Lecture: "Exploiting GPU'sfor scientific computing" (Leone Bosi, INFN Perugia)
20:30 Dinner
Dinner
20:30 - 20:30
Friday, 26 November 2010
08:30 Session 5
08:30 - 19:30
Contributions

08:30 Parallel programming theory - Tim Mattson (Intel)

09:20 An introduction to OpenMP - Tim Mattson (Intel)

10:30 Synchronization in OpenMP - Tim Mattson (Intel)

11:00 Work sharing constructs - Tim Mattson (Intel)

11:50 The OpenMP data environment - Tim Mattson (Intel)

14:30 OpenMP tasks - Tim Mattson (Intel)

15:15 OpenMP Memory model - Tim Mattson (Intel)

16:15 A survey of programming models - Tim Mattson (Intel)

18:30 Evening Lecture: "How to program a 1000 core processor" (Timothy G. Mattson, Intel Co.)
20:30 Dinner
Dinner
20:30 - 20:30
Saturday, 27 November 2010
08:30 Session 6
08:30 - 14:50
Contributions

08:30 Students feedback

09:00 Final examination

11:30 Delivery of certificates of attendance

14:30 Shuttle departure