With multidisciplinary expertise spanning technical pillars of high-performance computing, data science, and computational mathematics, we work toward building computational capabilities that position PNNL as a computing powerhouse. We also focus on enhancing the Science of Computing to achieve high-performance, power-efficient, and reliable computing at extreme scales for a spectrum of scientific endeavors that address significant problems of national interest, especially among PNNL’s core pursuitsenergy, the environment, national security, and fundamental science.
For his contribution to research recently showcased in a paper published in Nature, Tim Shippert, a scientist with the ARM Data Integration team in PNNL’s ACMD Division, ran six model years of ARM data to evaluate reviewers’ questions about specific calculations. The resulting work provided evidence of the first observed influence of atmospheric carbon dioxide at the Earth’s surface and helps confirm that climate models are accurately representing carbon dioxide’s impacts to climate.
The combination of leading-edge microscopy facilities, computational modeling, and federated data science capabilities—as well as cross-domain collaborations—can significantly advance fundamental scientific understanding and control of the synthesis and functionality of energy storage and conversion materials. A series of demonstrations, initially presented by scientists, including PNNL’s Kerstin Kleese van Dam, at SC14 in New Orleans, showcased the ongoing work of DOE’s Data Science Centers and how these collaborative, multi-institutional environments are improving methods for collecting, analyzing, and sharing Big Data and, most notably, driving innovation in materials science.
Bin Ren and Sriram Krishnamoorthy, both from PNNL’s High Performance Computing group, along with collaborators from Purdue University and Washington University in St. Louis, co-authored, “Efficient Execution of Recursive Programs on Commodity Vector Hardware,” which recently was accepted by the 36th Annual Association for Computing Machinery Special Interest Group on Programming Languages Conference on Programming Language Design and Implementation, known broadly as PLDI. Their paper, which presents a set of novel code transformations exposing data-parallelism in recursive, task-parallel programs that allows them to be mapped to commodity vector hardware, was among a record number of submissions with only 58 accepted for the conference.
As part of the upcoming Data Science Innovation Summit on Feb. 12-13, 2015 in San Diego, Calif., Kerstin Kleese van Dam, the Data Services Team Lead within PNNL’s ACMD Division, will present a keynote, “How to Use Streaming Data for Real-time Decision Making.” The summit will emphasize “Creating a Successful Data-Driven Culture” and feature thought leaders from industry, including corporate entities such as Google, Netflix, and Electronic Arts, as well as academic and government agencies, tackling topics related to advancing the role of data science in their respective business and research enterprises. The goal is to showcase how and why organizations are using Big Data and data analytics and the benefits of creating a data-driven culture.
For scientists in PNNL’s Advanced Computing, Mathematics, and Data Division, their work often crosscuts many domain science sectors within the Laboratory and among external collaborators. In this case, seeking methods to enhance data analytics of biological sequences using algorithmic graph theory led to a distinct intersection with work being done for high-performance computing applications contending with obstacles related to power constraints and massive data movement. For the scientists and their partners involved in this research, one point rings true: in science, the problems you start with may not be the only ones you solve.