In the last years, we assisted to an explosion of applications that require the elaboration of vast amounts of data. These applications search for hidden relations among data (big data analytics) or perform multidisciplinary computations to solve complex problems (scientific computing). These applications can unleash breakthrough innovation in society and industry but have extremely high computational demands. While heterogeneous platforms are well established for specific, embedded applications, they require substantial effort to optimize the execution of large, complex applications. Designers need to understand in depth the domain of the computation that often is complex and interdisciplinary, including Physics and Mathematics for scientific and machine learning applications, and to master hardware design for specializing the hardware platform and creating the proper data management infrastructure. This problem requires multidisciplinary skills and proper design environments.
This summer school aims at establishing a living and interdisciplinary environment to bridge the gap between system researchers and application developers. The former will learn most common application requirements, while the others will understand the potential of modern heterogeneous platforms and how to use them to solve their problems. The school will provide a bird’s-eye view on modern big data and scientific applications, highlighting their challenges and computational demands. Then, it will present the technology behind modern heterogeneous platforms from both industrial and research perspectives. Finally, it will discuss advanced design environments from bridging the gap between these two communities.