1 Answer

Background: What can the database community offer to HPC?
The relational model for databases revolutionized business data management since the 1970s, leading to today’s $30B market and a vibrant startup community. The hallmark of the approach is to “push the computation to the data,” insulating users and applications from details of data representation, scale, system architecture, and evaluation method, while affording runtime optimization opportunities unavailable to compile-time techniques. The features of this approach – a rigorous data model, a simple but expressive query algebra, cost-based optimization, declarative query languages, logical and physical data independence – have since been applied to new data types (trees, streams, graphs, arrays), new applications (financial markets, image analysis, DNA sequence analysis, social networking, finite element simulations) and new platforms (sensor networks, embedded systems, GPGPUs, shared-nothing commodity clusters, cloud platforms). This approach has only been minimally explored in the high-performance computing community, but current interest in data-intensive computing, simpler and accessible programming models, in situ analysis and visualization, and interactive control of long-running simulations offer new opportunities for engagement between the database community and HPC communities. In this workshop, we invite novel research that explores the design space between databases and HPC platforms.