GPU programming

<!--HTML--><h2>Lecture Contents</h2> <p>In this lecture, we will look into the basics of GPU computing to understand in which circumstances the usage of GPUs is beneficial for scientific computing. Using the Nvidia CUDA GPUs as examples,&nbsp;will learn how the hardware works, which guides us towards how it has to be programmed.</p> <h2><span style="color:#c0392b">Requirements</span></h2> <p>Students should have written basic C/C++ programs before, and should be familiar with pointers and arrays and/or vectors of data.</p> <p><strong>Hands on:&nbsp;</strong>We have the option to play with a few basic CUDA applications. There's three ways in which you can participate:</p> <ol> <li>Linux / Mac laptop with ssh.</li> <li>Any laptop and a browser.&nbsp;Please register on the current page&nbsp;to be granted access to GPUs on CERN's SWAN cluster. You will need a cernbox account, which is created once you visit <a href="https://cernbox.cern.ch">https://cernbox.cern.ch</a>. Note that the number of GPUs is limited.</li> <li>Team programming. If the two above are not for you, we will team up and work together.</li> </ol> <h3>Short Speaker Bio</h3> <p>Stephan obtained&nbsp;a PhD in particle physics, searching for decays of the Higgs boson with the ATLAS detector. Afterwards, he worked for the ROOT project at CERN, focussing on high-throughput computing&nbsp;and RooFit, a package for&nbsp;fitting and statistical analysis of data. Now, Stephan is a computing engineer in CERN IT's innovation group, focussing on GPU computing for high-energy physics.</p>