Lectures on Modern Scientific Programming

14 Nov 2016, 09:00 → 16 Nov 2016, 16:45 Europe/Budapest

Building 3, 2nd floor, Council Room (Wigner RCP)

Lectures on Modern Scientific Programming

The Wigner GPU-Lab organizes a three-day seminar series on Modern Scientific Programming between 14-16 November, 2016. These lectures review topics relevant to everyone working on scientific simulations or high-performance computing.
The lectures are highly recommended to those who are:

• Actively engaged in high-performance computing or analysis
• Author code where execution time is important
• Wish to learn (more) about parallel computing
• Would like to interface multiple languages and solve problems with rapid prototyping
• University students who plan to use programming daily in their field

The lectures may be of interest to those who are:

• Using interpreted math environments (Mathematica, Maple, Matlab, etc.) and wish to understand the run time costs
• Seasoned programmers but care to learn good practices and a few tricks

The lectures are free to attend, but registration is required for the secured entry to the KFKI Campus!

Monday, 14 November

09:00 → 12:00 Hardware and development tools

Modern Computing Hardware (CPU/GPU), Low-level insights to CPUs and GPUs, Overview of programming languages and paradigms

09:00 Overview of Modern Hardware Architectures and Technologies

What is in a microchip today? What metrics are important for certain applications? What technologies are arriving to the market in the near future?

10:15 Break / Questions and Answers

10:45 Platforms and Development tools

Overview of current events in the Windows-Linux world, introduction to some recommended Integrated Development Environments, debugging tools and other useful software.

12:00 → 13:30 Lunch break

13:30 → 16:45 Moving from C to C++

Integrated Development Environments, Version Control Systems, Build Systems

13:30 Gentle introduction to C++

Reasons for learning it, introduction to basic concepts and migration from C to C++

14:45 Break / Questions and Answers

15:15 Using C++ (Pitfalls, anti-patterns, libraries)

Compiling and linking issues, frequent mistakes, compiler errors, bad usage of language constructs and how to find and fix them. Some useful libraries for scientific applicaitons.

Tuesday, 15 November

09:00 → 12:00 Algorithms and Parallelism in C++

09:00 Algorithms and datastructures

10:15 Break / Questions and Answers

10:45 Parallel constructs and algorithms in C++

12:00 → 13:30 Lunch Break

13:30 → 16:30 Optimizing and Visualizing Simulations

13:30 Optimization usecases and pitfalls

14:45 Break / Questions and Answers

15:15 Visualization and Rendering

Wednesday, 16 November

09:00 → 12:00 Introduction to Python

09:00 Introduction to Python

10:15 Break / Questions and Answers

10:45 Python in Scientific Computing

12:00 → 13:30 Lunch Break

13:30 → 16:45 GPU Programming Today and Tomorrow

13:30 Portable GPGPU using SYCL and ComputeCpp

14:45 Break / Questions and Answers

15:15 News and trends in hardware, software, compilers and languages