I. Advanced Scientific Computing Research (ASCR)

The mission of the Advanced Scientific Computing Research (ASCR) Program is to deliver forefront computational and networking capabilities to extend the frontiers of science. A particular challenge of this program is fulfilling the science potential of emerging multi-core computing systems and other novel "extreme-scale" computing architectures, which will require significant modifications to today's tools and techniques.

Program Website: http://science.energy.gov/ascr

ASCR mission priorities:

  • To develop mathematical descriptions, models, methods, and algorithms to accurately describe and understand the behavior of complex systems involving processes that span vastly different time and/or length scales.
  • To develop the underlying understanding and software to make effective use of computers at extreme scales.
  • To transform extreme scale data from experiments and simulations into scientific insight.
  • To advance key areas of computational science and discovery that further advance the missions of the Office of Science through mutually beneficial partnerships.
  • To deliver the forefront computational and networking capabilities to extend the frontiers of science.
  • To develop networking and collaboration tools and facilities that enable scientists worldwide to work together.

The computing resources and high-speed networks required to meet Office of Science needs exceed the state-of-the-art by a significant margin. Furthermore, the algorithms, software tools, the software libraries and the distributed software environments needed to accelerate scientific discovery through modeling and simulation are beyond the realm of commercial interest. To establish and maintain DOE's modeling and simulation leadership in scientific areas that are important to its mission, ASCR operates Leadership Computing Facilities, a high-performance production computing center, and nation’s fastest high-speed network for science; and implements a broad base research portfolio in applied mathematics, computer science, computational science, and network research to solve complex problems on computational resources that are on a trajectory to reach well beyond a petascale within a few years.

The ASCR’s research areas of interest include:

(a) Applied Mathematics

This program supports fundamental research leading to mathematical advances and computational breakthroughs across DOE and Office of Science missions. Applied Mathematics research includes and supports efforts to develop robust mathematical models, algorithms and numerical software for enabling predictive scientific simulations of DOE-relevant complex systems. Important areas of supported research include: (1) novel numerical methods for the scalable solution of large-scale, linear and nonlinear systems of equations; (2) innovative approaches for analyzing and extracting insight from large-scale data sets; (3) efficient techniques for characterizing, propagating, and/or quantifying uncertainties and errors in next-generation solver, optimization, simulation, risk analysis, and other codes; (4) multiscale methods for continuous and/or discrete systems that efficiently account for physics and subcomponent interactions across vastly different time and length scales.

EXCLUSIONS: Development and/or implementation of existing numerical methods to a specific application are not within the scope of this program, no matter how challenging the application is.

(b) Computer Science

This program supports research that enables computing at extreme scales and the understanding of extreme scale data from both simulations and experiments. It aims to make scientific computers as easy and effective to use as possible. Research topics include advanced hardware and software architectures for exascale computing systems;  hardware and software approaches to power/energy management for HPC systems; programming models, languages, and compilers; execution models; scalable and fault tolerant operating and runtime systems, including file systems and input/output bottlenecks; programming environments and compilers; autotuning and performance modeling and assessment tools; software development tools and methods; scientific workflow systems;  scientific data management, integration, analysis and visualization for Petabyte to Exabyte data sets, both static and streaming, including in-situ methods. Research must have relevance to current and future high performance computing platforms as well as to the mission of the Office of Science. 

EXCLUSIONS: Quantum computing; networking; computer-supported collaboration; social computing; natural language processing/understanding/generation; generalized research in human-computer interaction; and research that is only applicable to hand-held, portable, desktop, cluster or cloud computing are not within the scope of this program.

(c) Network Environment Research

This program conducts system level research and development in next-generation networking and scientific collaborations to support diverse types of distributed science activities in the Office of Science.  Current areas of interest are 1) high-capacity optical networking technologies, and 2) data-intensive scientific collaborations.  High-capacity networking is focused on system level agile terabits networking that can deliver end-to-end throughput a thousand times (1000x) faster than today’s commercial Internet. Potential research topics for high-capacity networks include, but are not limited to, radically new architectures and protocols for hybrid terabits networking that support best-effort IP and dynamic circuits; composable and scalable transport protocols with end-to-end performance that far outstrip TCP/IP and its variants; storage and file systems extensions for 100 Gbps throughput and beyond.  Data-intensive scientific collaboration is focused on advanced collaboration services that enable distributed science teams to work together and share enterprise-level science facilities.  Potential topics of interest include, but are not limited to, end-to-end performance analysis tools and services; scientific workflows for optimizing collaborative resources; and distributed data management toolkits. Applicants pursuing graduate studies in either technical area are encouraged to use component-based approaches to facilitate the adoption of their proposed algorithms/software into DOE science infrastructures.

EXCLUSIONS: Sensornets, wireless networks, device level (switches, routers, etc.), photonics, and low level chip design and research activities are not within the scope of this program.