Registration Room: E3

Welcome to PDC and the Summer School

• Erwin Laure (PDC)
• Michael Hanke (KTH)

Room: E3

Concepts and Algorithms for Scientific Computing

• Björn Engquist (KTH and the University of Austin)

Room: E3

HPC Activities at Scania

• Mattias Chevalier (Scania)

Room: E3

Introduction to PDC's Environment

• Ul Hassan Izhar

Room: E3

Lab: Introduction to PDC's Environment

• Izhar Ul Hassan

Room: 4V2Röd, 4V3Ora

High-Performance Computer Architecture

• Erik Hagersten (Uppsala University)

Room: E3

High-Performance Computer Architecture

• Erik Hagersten (Uppsala University)

Room: E3

High-Performance Computer Architecture

• Erik Hagersten (Uppsala University)

Room: E3

Shared memory programming, OpenMP

• Thomas Ericsson (Chalmers University)

Room: E3 Two common ways of speeding up programs are to use processes or threads executing in parallel. The second week of the Summer School presents process-based parallel computing using MPI. In this lecture parallelization using threads will be discussed. Threads can be managed in different ways, e.g. using a low level library, like the POSIX threads library. The lecture deals with OpenMP, which is a more convenient way to use threads. OpenMP is a specification for a set of compiler directives, library routines, and environment variables that can be used to specify shared memory parallelism in Fortran and C/C++ programs. The threaded program can be executed on a shared memory computer (e.g. a multi- core processor). Typically, the iterations of a time consuming loop are shared among a team of threads, each thread working on part of the iterations. The loop is parallelized by writing a directive in the code and a compiler or a pre- processor will generate parallel code. Using short code examples, the lecture covers the most important OpenMP-constructs, how to use them and how not to use them. A few more realistic examples are presented as well.

Shared memory programming, OpenMP

• Thomas Ericsson (Chalmers University)

Room: E3 See preceding description.

Shared memory programming, OpenMP

• Thomas Ericsson (Chalmers University)

Room: E3 See preceding description.

Lab: Programming Exercises on OpenMP

• Niclas Jansson (CSC/NA)

Room: 4V2Röd, 4V3Ora

Lab: Programming Exercises on OpenMP

• Niclas Jansson (KTH)

Room: 4V2Röd, 4V3Ora

Code optimization

• Thomas Ericsson (Chalmers University)

Room: E3 In this lecture techniques for speeding up code on a uniprocessor are presented. The lecture covers basic optimization principles such as using the memory hierarchy in an efficient way. Using short code examples, the lecture covers stride minimization, blocking, loop- fusion, loop- splitting and other approaches. The choice of programming language (Fortran or C), and compiler will be discussed. There are some hints of how to tune Matlab programs. Other topics are vector arithmetic for floating point and elementary functions, virtual memory, files and the LAPACK and BLAS libraries.

Code optimization

• Thomas Ericsson (Chalmers University)

Room: E3 See preceding description.

Lab: OpenMP Advanced Project

• Niclas Jansson (CSC/NA)

Room: 4V2Röd, 4V3Ora

Lab: OpenMP Advanced Project

• Niclas Jansson (CSC/NA)

Room: 4V2Röd, 4V3Ora

Wrap up: OpenMP Advanced Project

• Niclas Jansson (CSC/NA)

Room: F3

GPU Architectures for Non-Graphics People

• David Black-Schaffer (Uppsala University)

Room: F3 David Black-Schaffer (Uppsala University) Today everyone is positioning GPUs for general purpose computing. They claim that you can get 10-100x speedups over conventional CPUs, and sometimes they're even right. However, to get the most out of current- (and next-) generation GPUs, one needs to understand the architectural differences and how they effect your choice of algorithm. In this talk I will cover GPU architecture in comparison to current CPUs, discuss the implications for getting good performance, and introduce OpenCL as a general-purpose programming language for accessing GPUs and CPUs today.

Introduction to OpenCL

• David Black-Schaffer (Uppsala University)

Room: F3

Lab: OpenCL

• David Black-Schaffer (Uppsala University)

Room: 4V2Röd, 4V3Ora

Lab: OpenCL

• David Black-Schaffer (Uppsala University)

Room: 4V2Röd, 4V3Ora

MPI Introduction

• Erwin Laure (PDC-HPC)

Room: E3

MPI Basic Concepts & Point-to-Point Communication

• Erwin Laure (PDC-HPC)

Room: E3

Lab: MPI Part 1

• Jonathan Vincent (PDC)

Room: 4V2Röd, 4V3Ora

Lab: MPI Part 1

• Jonathan Vincent (PDC)

Room: 4V2Röd, 4V3Ora

MPI - Collective Communication

• Erwin Laure (PDC-HPC)

Room: E3

MPI - Intermediate MPI

• Erwin Laure (PDC-HPC)

Room: E2

Performance Engineering

• Pekka Manninen (CSC, Finland)

Room: E2

Performance Engineering

• Pekka Manninen (CSC, Finland)

Room: E3

Lab: Performance Engineering

• Pekka Manninen (CSC, Finland)

Room: 4V2Röd, 4V3Ora

Wrap up: Performance Engineering

• Pekka Manninen (CSC, Finland)

Room: E3

MPI: Advanced Concepts

• Erwin Laure (PDC-HPC)

Room: E3

MPI: Advanced Concepts

• Erwin Laure (PDC-HPC)

Room: E3

Parallel Performance Engineering

• Pekka Manninen (CSC, Finland)

Room: E3

Lab: Parallel Performance Engineering

• Pekka Manninen (CSC, Finland)

Room: 4V2Röd, 4V3Ora

Lab: MPI Part 2

• Jonathan Vincent (PDC)

Room: 4V2Röd, 4V3Ora

Interconnection Networks

• Michael Schliephake (PDC)

Room: F3 Interconnect networks have large influence on the performance and efficiency of message-passing and File-I/O operations in cluster systems. The lecture presents concise information about interconnect networks in order to support the understanding of their influence and how to use it in favour of simulation applications.

Interconnection Networks

• Michael Schliephake (PDC)

Room: F3

Software co-design and the exascale challenge

• Mark Parsons (EPCC)

Room: F3 The next frontier of High Performance Computing is the exascale. By the early part of the 2020s we expect computers to exist that can perform 1018 calculations per second. In order to achieve this goal the systems will be incredibly complex with many millions of processor cores and complex memory and communication hierarchies. Programming such systems will be extremely challenging. This lecture will focus on the use of software co-design to prepare applications for exascale systems. It will explain how we need to use incremental and disruptive approaches to meet the many challenges posed by such systems and discuss some of the key issues we face over the next decade at the frontier of supercomputing.

Lab: MPI Part 3

• Jonathan Vincent (PDC)

Room: 4V2Röd, 4V3Ora

Future Programming Languages

• Stefano Markidis (PDC)

Room: E3 The lecture presents the current limitations of traditional programming models, such as MPI and openMP, and provides motivations to adopt new programming models for parallel computing. The potential of innovative programming models such as the Partitioned Global Address Space (PGAS) languages, and task-based programming models is discussed. Few examples of how to test on Lindgren these innovative programming models are also presented.

Future Programming Languages

• Stefano Markidis (PDC)

Room: E3 The lecture presents the current limitations of traditional programming models, such as MPI and openMP, and provides motivations to adopt new programming models for parallel computing. The potential of innovative programming models such as the Partitioned Global Address Space (PGAS) languages, and task-based programming models is discussed. Few examples of how to test on Lindgren these innovative programming models are also presented.

Abstractions for processing large data sets

• Jonas Yngvesson (Google)

Room: E3 At Google there is often a need to process very large data sets across many machines. Building efficient parallel processing programs is not trivial and to avoid the overhead of each engineer reinventing the wheel, Google has created several programming models to abstract away the complexities of parallelism and have the programmer concentrate on the core of his processing problem instead. I will talk mainly about the MapReduce model, how it works and the conceptual model the programmer has to work with when using it. I will also mention a bit about a more recent framework, Pregel, which is built for processing very large graph models.

Open Lab and Project Work Room: 4V2Röd, 4V3Ora