This GCoE is committed with the task of educating researchers and graduate students. We periodically organize courses and training activities.
The Programming and Tuning Massively Parallel Systems Summer School is aimed at enriching the skills of researchers, graduate students and teachers with cutting-edge technique and hands-on experience in developing applications for many-core processors with massively parallel computing resources like GPU accelerators.
BSC and UPC have been jointly organizing the Programming and Tuning Massively Parallel Systems (PUMPS) summer schools since 2010. The PUMPS summer school has received more than 200 applications each year, and has brought to Barcelona close to 100 students from all around Europe at each occasion. Professor Wen-mei W. Hwu (University of Illinois) and David Kirk (NVIDIA) have been the main instructors of PUMPS in both editions, although may other faculty members have been also involved: Nacho Navarro (UPC), David Kaeli (Northeastern University), Rosa M. Badia (BSC), Xavier Martorell (UPC), Jesus Labarta (UPC), Isaac Gelado (UPC and University of Illinois), and Juan Gomez-Luna (University of Córdoba).
The PUMPS summer school has evolved over time. The first edition had as main targets teaching the basics of CUDA to novices (beginners track), presenting hardware-specific optimizations to experimented programmers (advanced track), and teaching how to teach CUDA to faculty members (teachers track). Building on top of this experience, we have been focusing on advanced tracks and algorithmic optimizations. The teaching materials (lectures, notes, and laboratories) are continuously modified to focus on the new features included in CUDA. New lectures teach multi-GPU and CUDA/MPI programming patterns, two hot topics highly demanded by the attendants to the PUMPS summer school. Besides being a unique forum in Europe to learn about CUDA programming, the PUMPS summer school has also focused on bringing together CUDA developers from around Europe. Attendees can present their work at a poster session. A faculty committee awards the best poster with NVIDIA cards or CUDA books as a prize. Another set of donated equipment or books is the prize for the Best Achievement Award.
EuReCCA, the European Research Center on Computer Architecture, is the first HiPEAC EuroLab, being currently organized by Chalmers University of Technology (Sweden), Barcelona Supercomputing Center and UPC (Spain), and FORTH-ICS and Univ. of Crete (Greece) http://www.eurecca.eu/. We aim at consolidating the curriculum by offering joint degrees.
The aim of this course is to provide students with knowledge and hands-on experience in developing applications software for processors with massively parallel computing resources. The target audiences of the course are students who want to develop exciting applications for these processors, as well as those who want to develop programming tools and future implementations for these processors. We will be using NVIDIA processors and the CUDA programming tools in the lab section of the course.