Call for Participation
We are pleased to announce the 3rd DOE Performance, Portability and Productivity meeting; an opportunity for researchers to share ideas, practical experiences, and methodologies for tackling the compelling problems that lie at the intersection of performance, portability and productivity.
Authors are invited to submit novel research from all areas concerned with performance, portability and productivity. The meeting is anticipated to attract a diverse and cross-disciplinary audience from the DOE laboratories in addition to their industry and academic partners, bringing together: compiler, language and runtime experts; performance engineers; and domain scientists. Input from such a variety of experiences is critical to developing effective and productive solutions to performance portability, and tracking the highlights and lowlights of community experiences is key to identifying common themes and best practices.
Topics of interest include:
- Extensions to standard languages, libraries and runtimes such as C/C++, Fortran, OpenMP, OpenACC, SYCL, OpenCL
- Algorithmic and application development techniques
- Software tools, libraries, domain specific languages and other abstractions
- Case studies, with both positive and negative experiences, documenting efforts to run across multiple diverse platforms using state-of-the-art tools and techniques
- Achieving performance portability for legacy codes
- Preparing applications for unexpected changes in architecture, e.g. managing multiple memory hierarchies
- Definitions of and metrics for measuring performance portability
- Experiences with burst buffers and other I/O approaches
- Productivity concerns related to the above topics
Note that attendance by parties outside of the DOE labs, the lab collaborators, and the vendors involved in current COEs will be limited and by invitation only. If you did not explicitly receive an invitation, please submit your request to a steering committee member.
The meeting organizers are seeking proposals for:
- Technical presentations (15 or 30 minutes)
- Breakout session topics (1.5 hours)
- Panels on a focus topic area (1 hour)
- And if you just want to be an active participant, you're welcome to register
The work can be something you've already presented or published at a prior conference or workshop. It can be a work in progress with preliminary results and observations, in fact this is highly encouraged as it will generate good discussion and feedback for your work.
Submissions are in the form of an abstract, 1 to 2 paragraphs. The submission form is limited to text only, i.e. no attachments, but feel free to include links to publications and/or illustrations that support your submission.
Submissions due: January 23rd, 2019
- Notifications: February
- Hotel room block closes: March 1st, 2019
- Meeting: April 2nd, 3rd and 4th (1/2 day on the 4th)
The following will be required of all participants:
- Refrain from discussing information held under non-disclosure agreements. Contact a steering committee representative if you need specific guidance.
- In the spirit of the meeting, talks and discussions should address general challenges to the goal of performance portability and approaches that might be applied to overcome those challenges.
- The focus will be on portable, non-vendor-specific solutions as seen from the application developer perspective (that is, abstractions that hide vendor-specific solutions are acceptable). It is expected that a particular focus of the meeting will be to address possible evolutions of current standards (for example, OpenMP and C++) to better support performance portability.
- Projections to future machines should not be presented.
- Talks and discussions must be unclassified and non-sensitive in nature.
- Participants must accept that the DOE will have multiple target platforms as part of their national strategy and you are joining the discussion in the spirit of cooperation. All COEs are working toward the goal of making these platforms the most useful and high performance they can be without the threat of "vendor lock-in."