About

Co-­design  refers  to  a  computer  system  design  process  where  scientific  problem requirements influence architecture design and technology and constraints inform formulation   and   design   of   algorithms   and   software,   while   co-­‐optimization   refers   to optimization of hardware, software, and algorithms for a given application. Consistent with DOE’s effort, we need to start  formulating  efforts  in  scoping  the  different  applications  and the parameters of architecture and in  mapping  them  with  each  other.  We  think  that  given the increasing need for new materials in multiple applications, major ongoing research and development centers of computational and physical sciences need to be formally engaged in the hardware, software, numerical methods, algorithms, and applications co-­‐design process. As articulated by DOE, co-­‐design methodology requires the combined expertise of vendors, hardware architects, system software developers, domain scientists  such  as  material  scientists, biologists, chemists, and condensed matter physicists, computer scientists, and applied mathematicians working together to make informed decisions about features and trade-­‐offs in the design of the hardware, software and underlying algorithms.

In  this  meeting,  we  intend  to  discuss  Co-­‐design  3.0,  where  a  more  adaptable  and scalable paradigm in which systems can be dynamically configured for different applications.    Co-­‐design  refers  to  a  computer  system  design  process  where  application requirements are collocated with system and processor architectures, design, algorithms and software. Extreme Computing refers to the limits of computing given the constraints, for solving real problems.

The premise is that with the slowing of Moore’s law and the need for computing to solve problems for addressing societal needs, we need to focus on real applications as they evolve in time compared to standard benchmarks. For this to be practically viable, this should be done in a scalable framework for lower cost. We think that major ongoing research and development centers of computational and physical sciences need to be formally  engaged  in  the  co-­‐design  of  hardware,  software,  numerical  methods,  algorithms, and applications.

The  all-invited  meeting  is  expected  to  include  renowned  experts  from  academia, Department of Energy, National Institutes of Health, and industry discuss aspects of  architecture, hardware, software, and applications as a variety of grand challenge problems in    materials    science,    chemistry,    systems    biology,    bio-­‐informatics,    data    analysis, visualization, machine learning etc. Our intent is to develop a wider research, education, and practice  for computing  as an  enabler for solution  to major societal  problems.