IDEAS-Classic Team

The IDEAS-Classic Project Team

The IDEAS-Classic Project (ASCR/BER partnership) has members from seven US DOE laboratories. Efforts are organized into four focus areas:

Use Cases (lead: J. David Moulton): IDEAS has three concrete use cases that serves as focal points for our productivity efforts. IDEAS has two dual goals: Satisfy the requirements of our use cases and provide productivity capabilities to the broader community through meeting use case requirements.
Extreme-scale Scientific Software Development Kit (xSDK) (lead: Lois Curfman McInnes): The IDEAS xSDK is a collection of independently-developed, reusable scientific software components and libraries (currently the numerical libraries hypre, PETSc, SuperLU and Trilinos and the biogeochemistry component Alquimia), along with interoperability layers, build tools, testing, and documentation that enable xSDK users to employ these capabilities as a seamless software environment.
Methodologies for Software Productivity (How-To) (lead: Michael Heroux): This effort is focused on observing, capturing and promoting best practices in scientific software engineering. Also known as the “how to” focus area, this effort is intent on creating content that will help the IDEAS project be more productive and help the community at large by producing high value content on topics such as how to start an open source scientific software project, how to select source management, issue tracking and other important tools.
Outreach and Community (lead: David Bernholdt): This effort is focused on establishing connections to the broader community, which is essential in order to leverage the value of the content the IDEAS Project develops and complement external efforts. The IDEAS Project succeeds by satisfying the specific requirements of our use cases and by disseminating our work to the broader scientific software community. Furthermore, our efforts are part of a broader collection of related projects.

The structure of the multi-institutional IDEAS project is shown in the diagram above. The team is listed below the diagram by project category.

Methodologies for Software Productivity

Lead: Michael Heroux (SNL)

Adapt software engineering practices for extreme-scale science, with emphasis on workflows, lifecycles, and goals/questions/metrics.

Ross Bartlett (ORNL)

Mark Berrill (ORNL)

Anshu Dubey (ANL)

Todd Gamblin* (LLNL)

Osni Marques (LBNL)

Pat McCormick (LANL)

Nicholas Moss (LANL)

Andrew Salinger* (SNL)

Jason Sarich (ANL)

Dali Wang (ORNL)

Extreme-Scale Scientific Software Development Kit

Lead: Lois Curfman McInnes (ANL)

Improve interoperability and compatibility across xSDK libraries; address performance portability on new and emerging architectures; develop and extract scientific domain components, including coupling functionality for multiphysics, multiscale models.