iCSC2008  Towards Reconfigurable High-Performance Computing

Lecturers:
Iris Christadler - Leibniz Supercomputing Centre - Germany
Manfred Muecke - University of Vienna - Austria
Andrzej Nowak  - CERN, Geneva

A few questions

  • What will tomorrow's supercomputers look like?

  • What are the expected changes in commodity PC hardware and why should we care about them?

  • What should I do to use tomorrow’s supercomputers efficiently?

  • How to port your application to GPUs (without porting it)?.

  • Why thinking parallel will make all the difference?

  • Shall we think in local or global address spaces?

  • Do we need data stream processing to cope efficiently with many-core CPUs?

  • What to think of the new programming languages … Fortress, Chapel, X10, UPC, Co-Array Fortran?

  • Do you need to learn VHDL when using FPGAs?

  • How to define an FPGA's peak performance? (and how to cheat doing so?)

  • How can FPGAs running at 100MHz outperform CPUs running at 3GHz

  • Does C-to-Hardware work?

  • Have you ever wondered what the acronyms DEISA/PRACE/HPCS mean?

  • Can your playstation save the world?

  • Why should supercomputers care about the climate change?

  • Do you still believe that Roadrunner is just a bird and Maxwell is a Scottish physicist?

 

All the answers in the Computational Intelligence at iCSC

Moore's law still holds and provides us with unprecedented device integration, resulting in abundant logic resources even on commodity computing platforms. However, computing has failed to take advantage of this gift and increase in computing performance is constantly lagging behind the increase in logic resources.
 

In short: semiconductor technology has overtaken chip designers, computer scientists and programmers on the right. This is strongly felt in high-performance computing (HPC) where most applications can no longer take advantage of the many cores provided by current supercomputers. Stalling cores nevertheless take up energy and in HPC power consumption is becoming an important issue. Among the more promising future solutions to these problems is reconfigurable computing (computing on flexible fabrics). In this series of lectures, we want to explore the reasoning behind reconfigurable HPC, its prospects, implications and issues.

We will discuss different hardware architectures and their respective requirements and implications to the model of computation applied (or the mismatch thereof). We will compare them, sketch their potential for HPC and introduce a more unified view on them.

We will show that the dominant problem in HPC is not hardware but software. Especially the fact that our programming models do not match current technology is the root of many inefficiencies.

Reconfigurable Computing is challenging, because it asks many questions at the same time. But it is also most rewarding, because it forces us to rethink the way we design computers, interconnects, compilers and applications.

The lectures aim at qualifying students to understand where and why reconfigurable computing can be expected to have a considerable impact on tomorrows high-performance computing landscape, and where not.
 

Overview

Slot

Lecture

Description

Lecturer

   

Monday 3 March

 

10:15-

10:30

Introduction

School opening -

 

10:30- 11:25

Lecture 1

Basics

Iris Christadler

11:35- 12:30

Lecture 2

Multicore Architectures

Andrzej Nowak 

12:30- 14:00

 

Lunch

 

14:00- 14:55

Lecture 3

Platforms I: Advanced Architectural Features

Andrzej Nowak

15:05- 16:00

Lecture 4

Platforms II: Special-Purpose Accelerators

Andrzej Nowak

16:00- 16:30

 

Coffee break

 

16:30- 17:25

Lecture 5

Multicores at work: The CELL Processor

Iris Christadler

17:30

 

Adjourn

 
   

Tuesday 4 March

 

09:00-

09:55

Lecture 6

Platforms III - Programmable Logic

Manfred Muecke

10:05- 11:00

Lecture 7

Reconfigurable HPC I - Introduction

Iris Christadler

11:00- 11:30

 

Coffee break

 

11:30- 12:25

Lecture 8

Reconfigurable HPC II - HW Design Methodology,
Theory & Tools

Manfred Muecke

12:30- 14:00

 

Lunch

 

14:00- 14:55

Lecture 9

Advanced and Emerging Parallel Programming Paradigms

Manfred Muecke

15:05- 16:00

Lecture 10

Summary: Hybrid Platforms, Hybrid Programming?

Iris Christadler

16:10- 16:25

Theme closing

Transition between HPC and Data Analysis themes: Using HPC concepts in data analysis software (short session)

Alfio Lazzaro

16:30

 

Adjourn

 

Detailed description of all lecturers