Ninth INFN International School on: "Architectures, tools and methodologies for developing efficient large scale scientific computing applications" ESC17 - Bertinoro (Forlì-Cesena) Italy 22-28 October 2017

22 Oct 2017, 16:00 → 28 Oct 2017, 15:00 Europe/Rome

Bertinoro

Hands-on material https://infn-esc.github.io/esc17/

Pictures https://www.flickr.com/gp/felicepantaleo/E2V710 https://www.flickr.com/photos/domenico-galli-bologna/albums/72157686632463622

Sunday 22 October

Registration and Welcome

20:30 Welcome Dinner

Monday 23 October

Session 1

1 Welcome and introduction

Speaker: Mauro Morandin (INFN - Padova)

Slides welcome17.pdf

2 Computer Architecture evolution and the performance challenge

Speaker: Vincenzo Innocente (CERN)

Slides Architecture_ESC16_byVI.pdf

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 Efficient C++ programming and memory management

Speaker: Dr Francesco Giacomini (CNAF)

Slides cpp-esc17.pdf

6 Efficient C++ programming and memory management

Speaker: Dr Francesco Giacomini (CNAF)

15:30 Coffee break

7 Efficient C++ programming and memory management

Speaker: Dr Francesco Giacomini (CNAF)

8 Consolidation

9 Students lightning presentations

Slides Monday_presentations.pdf

19:30 Dinner

Tuesday 24 October

Session 2

10 Efficient C++ programming and memory management

Speaker: Dr Francesco Giacomini (CNAF)

11 Efficient C++ programming and memory management

Speaker: Dr Francesco Giacomini (CNAF)

10:00 Coffee break

12 Efficient C++ programming and memory management

Speaker: Francesco Giacomini (CNAF)

13 Efficient C++ programming and memory management

Speaker: Francesco Giacomini (CNAF)

14 Consolidation

13:30 Lunch break

15 Introduction to parallel computing (basic concepts)

Speaker: Dr Tim Mattson (Intel)

Slides intro_par_prog_with_openMP_rotated.pdf

16 Introduction to parallel computing with OpenMP

Speaker: Dr Tim Mattson (Intel)

16:30 Coffee break

17 Parallel Performance concepts using OpenMP

Speaker: Dr Tim Mattson (Intel)

18 Students lightning presentations

Slides Tuesday_presentations.pdf

19:30 Dinner

Wednesday 25 October

Session 3

19 Efficient C++ programming and memory management

Speaker: Francesco Giacomini (CNAF)

Slides memory-esc17.pdf

20 Efficient C++ programming and memory management

Speaker: Francesco Giacomini (CNAF)

10:00 Coffee break

21 A "Hands-on" introduction to OpenMP

Speaker: Dr Tim Mattson (Intel)

22 A "Hands-on" introduction to OpenMP

Speaker: Dr Tim Mattson (Intel)

23 Consolidation

13:00 Lunch break

14:00 Visit of the Interfaith Museum

24 Working with OpenMP: Performance Optimization

Speaker: Dr Tim Mattson (Intel)

16:15 Coffee break

25 Working with OpenMP: Debugging Applications

Speaker: Dr Tim Mattson (Intel)

26 Consolidation

19:30 Dinner

Thursday 26 October

Session 4

27 Efficient floating-point computation and vectorization

Speaker: Vincenzo Innocente (CERN)

Slides Vincenzo_OptimalFloatingPoint2017.pdf

28 Efficient floating-point computation and vectorization

Speaker: Vincenzo Innocente (CERN)

10:00 Coffee break

29 Efficient floating-point computation and vectorization

Speaker: Vincenzo Innocente (CERN)

1 https://infn-esc.github.io/esc17/vectorization/binning.html

30 Consolidation

31 GPUs and the Heterogeneous programming problem

Speaker: Dr Tim Mattson (Intel)

Slides GPGPU_intro.pdf

13:00 Lunch

32 GPU programming with OpenCL: Ideas and the host program

Speaker: Dr Tim Mattson (Intel)

Slides OpenCL_programming.pdf

33 Consolidation

16:15 Coffee break

34 Consolidation

35 From HPC to Deep Learning, a short journey on modern techniques

Software design of legacy high-performance computing (HPC) applications is heavily influenced by the frequent changes on the architectures. The classic CPU-based approach with communication libraries is not sufficient to scale beyond the limit of Amdahl law and Dennard scaling. Silicon manufacturers design alternative architectures and programming models to overcome this limitation, leading to systems with hidden complexity. In this seminar are presented a variety of possible scalable configurations and details of usability of each one, with specific attention to NVIDIA software tools. A second part of the talk is dedicated to the intriguing emerging technique of deep learning (DL). Two aspects of DL are shown in detail: 1) it is a new technology which allows to create fast-simulation software, 2) DL changes completely the paradigm of programming, moving the developer to a much higher abstraction layer, away from the hardware details.

Speaker: Piero Altoè (NVIDIA)

20:00 Social dinner

Friday 27 October

Sessione di venerdi'

36 Programming GPUs with OpenCL: Kernel programs

Speaker: Dr Tim Mattson (Intel)

37 Using GPUs with other programming models

Speaker: Dr Tim Mattson (Intel)

Slides par4_Other_GPGPU_prog_envs.pdf

10:00 Coffee break

38 GPGPU epilog

Slides par5_GPGPU_epilog.pdf

39 Cluster Computing with MPI

Speaker: Dr Tim Mattson (Intel)

Slides MPI_programming.pdf

40 Consolidation

13:00 Lunch

41 The 10 core constructs every MPI programmer should know

Speaker: Dr Tim Mattson (Intel)

42 Geometric decomposition and MPI

Speaker: Dr Tim Mattson (Intel)

16:30 Coffee break

43 Information

Speaker: Mauro Morandin (PD)

44 Consolidation

45 Surviving the Red Queen’s Race: A guide for the perplexed programmer

The Red Queen’s race is an important hypothesis in evolutionary biology raised to explain the impact of competition on the rate of evolution. I submit that it is an excellent analogy for the hardware chaos we are now facing. The question is, what can programmers do to survive as the hardware literally changes underneath them? In this talk we will look at trends in hardware and explore some of the approaches being used by software developers to cope.

Speaker: Dr Tim Mattson (Intel)

Slides Red_queens_race_survival.pdf

19:30 Guitar Cocktail

20:30 Dinner

Saturday 28 October

Session 9

46 Students feedback

47 Final examination

11:00 Coffee break

48 Delivery of certificates of attendance

12:00 Lunch

49 Shuttle departure (to Forli' railway station)