Eighth INFN International School on: "Architectures, tools and methodologies for developing efficient large scale scientific computing applications" ESC16 - Bertinoro (Forlì-Cesena) Italy 23-29 October 2016

23 Oct 2016, 16:00 → 29 Oct 2016, 15:00 Europe/Rome

Bertinoro

Hands-on information https://infn-esc.github.io/esc16/

Sunday, 23 October

Registration and Welcome

20:30 Welcome Dinner

Monday, 24 October

Session 1

1 Welcome and introduction

Speaker: Mauro Morandin (INFN - Padova)

2 Computer Architecture evolution and the performance challenge

Speaker: Vincenzo Innocente (CERN)

10:30 Coffee break

3 Computer Architecture evolution and the performance challenge

Speaker: Vincenzo Innocente (CERN)

4 Hands-on environment checkout

Speaker: Dr Francesco Giacomini (CNAF)

12:30 Lunch break

5 Welcome by the Director of the INFN Padova site

Speaker: Mauro Mezzetto (PD)

6 Efficient C++ programming and memory management

Speaker: Dr Francesco Giacomini (CNAF)

7 Efficient C++ programming and memory management

Speaker: Dr Francesco Giacomini (CNAF)

15:40 Coffee break

8 Efficient C++ programming and memory management

Speaker: Dr Francesco Giacomini (CNAF)

9 Consolidation

10 Students lightning presentations

20:30 Dinner

Tuesday, 25 October

Session 2

11 Efficient C++ programming and memory management

Speaker: Dr Francesco Giacomini (CNAF)

12 Efficient C++ programming and memory management

Speaker: Dr Francesco Giacomini (CNAF)

10:00 Coffee break

13 Efficient C++ programming and memory management

Speaker: Francesco Giacomini (CNAF)

14 Efficient C++ programming and memory management

Speaker: Francesco Giacomini (CNAF)

15 Consolidation

13:30 Lunch break

16 Introduction to parallel computing (basic concepts)

Speaker: Dr Tim Mattson (Intel)

17 Introduction to parallel computing with OpenMP

Speaker: Dr Tim Mattson (Intel)

16:30 Coffee break

18 Parallel Performance concepts using OpenMP

Speaker: Dr Tim Mattson (Intel)

19 Students lightning presentations

20:30 Dinner

Wednesday, 26 October

Session 3

20 Efficient C++ programming and memory management

Speaker: Francesco Giacomini (CNAF)

21 Efficient C++ programming and memory management

Speaker: Francesco Giacomini (CNAF)

10:00 Coffee break

22 A "Hands-on" introduction to OpenMP

Speaker: Dr Tim Mattson (Intel)

23 A "Hands-on" introduction to OpenMP

Speaker: Dr Tim Mattson (Intel)

24 Consolidation

13:00 Lunch break

25 Working with OpenMP: Performance Optimization

Speaker: Dr Tim Mattson (Intel)

26 Working with OpenMP: Debugging Applications

Speaker: Dr Tim Mattson (Intel)

16:30 Coffee break

27 Consolidation

28 Evening Lecture "Go: a different tack at building concurrent programs that grow with grace"

In this talk, we'll introduce the basic concepts of Go, focusing on its concurrency primitives and its interface model. We'll see how these concepts, together with a great tooling and quick edit-compile-run cycle allow for a great development environment. Finally, we'll touch on how these assets apply to science software: from control frameworks to soft real-time data acquisition systems.

Speaker: Dr Sebastien Binet (LPC/IN2P3)

19:30 Dinner

Thursday, 27 October

Session 4

29 Efficient floating-point computation and vectorization

Speaker: Vincenzo Innocente (CERN)

30 Efficient floating-point computation and vectorization

Speaker: Vincenzo Innocente (CERN)

10:00 Coffee break

31 Efficient floating-point computation and vectorization

Speaker: Vincenzo Innocente (CERN)

32 Consolidation

33 GPUs and the Heterogeneous programming problem

Speaker: Dr Tim Mattson (Intel)

13:00 Lunch

14:30 Visit of the Interfaith Museum

34 GPU programming with OpenCL: Ideas and the host program

Speaker: Dr Tim Mattson (Intel)

35 Consolidation

17:15 Coffee break

36 Consolidation

20:00 Social dinner

Friday, 28 October

Sessione di venerdi'

37 Programming GPUs with OpenCL: Kernel programs

Speaker: Dr Tim Mattson (Intel)

38 Programming GPUs with OpenCL: Performance issues

Speaker: Dr Tim Mattson (Intel)

10:00 Coffee break

39 Lecture

40 Cluster Computing with MPI

Speaker: Dr Tim Mattson (Intel)

41 Consolidation

13:00 Lunch

42 The 10 core constructs every MPI programmer should know

Speaker: Dr Tim Mattson (Intel)

43 Geometric decomposition and MPI

Speaker: Dr Tim Mattson (Intel)

16:30 Coffee break

44 Information

Speaker: Mauro Morandin (PD)

45 Consolidation

46 Evening lecture: The future of Big Data: Polystore, specialized storage engines, and embedded analytics.

Theory is nice when trying to understand Big Data systems, but nothing beats experience with real data. Working with the MIMIC II data set (data from an intensive care unit) we've concluded that: 1. Data must match the storage engine if you care about performance 2. Data in flat files is almost equivalent to deleting it. Or, turning these conclusions into slogans, "one size does not fit all" and "we need to bring the power of a database to all data". In this talk we describe our ongoing work to create a system that responds to these slogans. We call this the BigDAWG Polystore system. A Polystore system contains multiple storage engines integrated behind a common API but exposing features of individual storage engines as needed. We are also working on a new storage engine tuned to the needs of sparse array data called TileDB. TileDB has entered production usage at the Broad Genomics institute. Our continuing work with TileDB is to extend it to dense arrays (thereby competing with HDF5). Finally, we believe that key analytics functions need to be integrated into the storage engines. We'll describe our early efforts to create GraphBLAS routines and other machine learning primitives integrated into our Polystore system.

Speaker: Dr Tim Mattson (Intel)

19:30 Dinner

Saturday, 29 October

Session 9

47 Students feedback

48 Final examination

11:00 Coffee break

49 Delivery of certificates of attendance

12:00 Lunch

50 Shuttle departure (to Forli' railway station)