Introduction to GPGPU and CUDA programming

This course gives an overview on the GPUs .These devices, originally designed for graphics applications, accelerate the calculation of the computational core of HPC applications.

The attendees will be introduced to the basic concepts of the language, the main features of this programming model and to the performance tuning. The course is mainly focused on the CUDA (Compute Unified Device Architecture) programming language, but some introductory notes on OpenCL, OpenACC as well as brief overview on other types of accelerators (Intel MIC) will be given.

Course Organizer:

P. Dagna (Milan), G. Fatigati (Bologna), L. Ferraro (Rome)

3 Editions in Italian:

29-30 January: Rome in Italian - SLIDE
9-10 May: Bologna in English. PATC course
7-8 November: Milan in Italian

Click on the date to register

Topics:

GPU NVIDIA architecture, CUDA programming and execution model, CUDA libraries, performance tuning, new Kepler GPU features, introductory notes on OpenCL, OpenACC, Intel MIC.

Target audience:

Researchers and programmers interested in scientific programming and efficient use of GPGPU in HPC context.

Pre-requisites:

A basic knowledge on C programming and Linux or Unix. A basic knowledge of any parallel programming is reccomended.

Agenda January Course

Duration:

2 dd

Contacts:

For further information about HPC courses, contact corsicineca.it.

Courses are hosted in CINECA premises:

Cineca - BOLOGNA, Via Magnanelli 6/3, 40033 Casalecchio di Reno, Bologna ph. + 39 051 6171411
Cineca - MILANO, Via R. Sanzio 4, 20090 Segrate, Milan ph. +39 02 269951
Cineca - ROMA, Via dei Tizii, 6 - 00185 Rome ph. +39 06 444861

Credits | Home