Motivation

Informed by the SC2011 theme of data-intensive computing, we propose to bring together researchers in the databases/data management community (exemplified by venues such as SIGMOD, VLDB, and ICDE) and the high performance computing community to identify opportunities for collaboration and set new research directions.

Our thesis is that the techniques and strategies developed in the database community – algebraic optimization, simple data parallel languages, data models as first-class citizens, logical data independence – are applicable to problems faced by the HPC community as we enter the exascale age. This focus complements the Petascale Data Storage Workshop series that has been collocated with SC for the past five years, since their focus has been solely on storage for HPC as opposed to application of data modeling, language, optimization, and indexing techniques pioneered by the database community.

By associating the workshop at SC2011 as opposed to a database conference, we are asserting that the applications and platforms considered by the HPC community should drive the focus of the workshop. We seek new approaches to the problems in these areas as opposed to providing yet another forum for the large and growing body of work in “big data.” In particular, we request that each position paper include a section considering how the work could be deployed in the context of leadership-class computing platforms and/or address applications of interest in HPC.

We focus our advertising on attracting database researchers to participate in a forum outside their “comfort zone.” Our strategy is to invite key players who have demonstrable perspective on both the HPC and database research landscapes, advertise to key mailing lists (DBWorld), and reach out to specific individuals and organizations who we believe will be enthusiastic about the topic. A list of proposed panel members and participants appears below.