Architectural trends indicate two directions. One being, heterogenous system equipped with accelerators for example the current system – Summit and the up and coming exascale systems such as Frontier and El Capitan and the other being homogeneous arm-based systems such as GW4 Isambard or the next Isambard 2 and Fugaku. A heavy reliance of future exascale machines for performance still seems to be ones with accelerators. One could even argue that the vector units of the curent HPC Arm processors are accelerators themselves. We are yet to see what will top the chart in a few years from now.
The ever-increasing heterogeneity in supercomputing applications has given rise to complex compute node architectures offering multiple, heterogeneous levels of massive parallelism. As a result, the ‘X’ in MPI+X demands more focus. Exploiting the maximum parallelism out of such systems necessitates sophisticated programming approaches that can provide scalable as well as portable solutions without compromising on performance. While an obvious expectation from the a programmer is to deliver portable solutions, thus avoiding duplicate programming effort, it is also important to understand the critical assessments of the possibilities and limitations of minimal-maintenance/performance portability approaches.
Literature shows that the abstraction can be raised at different levels i.e. at the high-level using directives such as OpenMP and OpenACC or at the language level using frameworks such as Kokkos and Raja or at a lower level using OpenCL and/or OneAPI. All of these efforts have certainly led to better compiler technologies.
Toward this end, the workshop will highlight transformational work along with improvements over state-of-art through the accepted papers and prompt discussion through keynote/panel that draws the community’s attention to key areas that will facilitate the transition to accelerator-based high-performance computing (HPC). The workshop aims to showcase all aspects of heterogeneous systems discussing innovative high-level language features, lessons learned while using directives to migrate scientific legacy code to parallel processors, compilation and runtime scheduling techniques among others.
WACCPD2020 will be co-located with SC20, Atlanta. In the past six years of this workshop, WACCPD has been one of the major forums at SC to bring together programming model users, developers, and tools community to share knowledge and experiences to tackle emerging complex parallel computing systems.
Topics of interest for workshop submissions include (but are not limited to)
- Application developers’ experiences porting scientific applications to modern systems
- Critical assessment of the possibilities and limitations of minimal-maintenance/performance and portability approaches – Support the analysis with case studies
- Compiler and runtime support for current and emerging architectures (e.g. heterogeneous architectures, low-power processors)
- Prototypes of behaviors of parallelism constructs in the base language while used within accelerator routines
- Design and development of directives created for complex parallel patterns
- Abstract handling of complex/heterogeneous memory hierarchies
- Extensions to and shortcomings of current directives for heterogeneous systems
- Compare and contrast high-level and lower-level abstractions with respect to performance, tuning, programmer productivity and overall analysis
- Modeling, verification and performance analysis tools
- Auto-tuning and optimization strategies
- Extending directive-based approaches to other environments such as Julia or Python
- Parallel computing using hybrid programming paradigms (e.g. MPI, OpenMP/OpenMP offloading,
- OpenACC, OpenSHMEM, OneAPI)
- Asynchronous execution and scheduling (task-based approaches)
- Scientific libraries interoperability with directive-based models
- Power/energy studies and solutions targeting accelerators or heterogeneous systems
Workshop Important Deadlines
Paper Submission Deadline:
August 24, 2020 AOEAugust 28, 2020 AOE (Extended, firm deadline)
- Author Notification: October 5, 2020
- Workshop Ready Deadline: October 19, 2020 AOE
- Camera Ready Deadline: December 4, 2020 AOE (Note: This deadline is after your presentation)
Best Paper Award
The Best Paper Award will be selected on the basis of explicit recommendations of the reviewers and their scoring towards the paper’s originality and quality. In order to be considered for the Best Paper Award, the authors must submit an Artifact Description appendix according to the reproducibility initiative of the SC20 technical papers.
WACCPD papers will be peer-reviewed and selected for presentation at the workshop. The paper presented will be published as post-proceedings in Lecture Notes in Computer Science (LNCS) with Springer. The camera-ready version will be due after the workshop. The authors are expected to use this additional time for further refining their manuscript based on the feedback received from the workshop along with reviews from the peer-review process.
For publication in LNCS, the Springer guidelines must be followed:
Authors should consult Springer’s authors’ guidelines and use their proceedings templates, either for LaTeX or for Word, for the preparation of their papers. Springer encourages authors to include their ORCIDs in their papers. In addition, the corresponding author of each paper, acting on behalf of all of the authors of that paper, must complete and sign a Consent-to-Publish form. The corresponding author signing the copyright form should match the corresponding author marked on the paper. Once the files have been sent to Springer, changes relating to the authorship of the papers cannot be made.
Authors can use the Springer LNCS template provided on Overleaf if they want to prepare their submission using that service.
Paper Submission Guidelines
- Papers should be submitted electronically via the SC20 Submission Page.
- They must follow the Springer LNCS format (LaTeX or Word) and follow Springer’s authors’ guidelines (Note: This is an FTP link).
- Submissions are limited to 18 pages. The 18-page limit includes figures, tables, and appendices, but does not include references, for which there is no page limit. It also does not include pages for describing data artifacts and reproducibility actions as outlined in the reproducibility initiative of this workshop.
- Submitted papers should not have appeared in or be under consideration for a different workshop, conference or journal.
- In submitting the paper, the authors acknowledge that at least one author of an accepted submission will register for and attend the workshop.
- Camera ready submissions can have a total of 20 pages (excluding references and artifacts/reproducibility information) to accommodate feedback from the workshop.
WACCPD supports SC20’s reproducibility initiative. In contrast to the SC20 technical paper submission process, WACCPD is not enforcing reproducibility information, however, strongly encourages submitters to provide artifact appendices in form of an Artifact Description (AD) and eventually an additional Artifact Evaluation (AE).
Authors should follow the instructions on SC20’s reproducibility web page. Similar to SC20’s reproducibility process, AD/AE appendices will be automatically-generated from an AD/AE appendix form provided on the WACCPD submission page. Authors are encouraged to familiarize themselves early with the AD/AE appendix requirements on the SC20 web page and the WACCPD’s sample submission form.
From the artifact information inserted to the submission form, Linklings will automatically generate LaTeX source files compliant with the LNCS format. If authors choose to publish this information with Springer, these generated LaTeX files will be part of the submission. There is (currently) no page limit for AD/AE appendices.
Springer will provide an “Artifacts Available” badge (or similar) to put on qualified papers. Springer’s exact requirements for this badge are under evaluation.
Workshop Reproducibility Guidelines
- A paper cannot be disqualified based on information provided or not provided in this appendix, nor if the appendix is not available.
- The artifact information will be available to reviewers.
- The availability and quality of an appendix can be used in ranking a paper. In particular, if two papers are of similar quality, the existence and quality of the appendices can be part of the evaluation process.
- For providing data and corresponding global object identifiers, authors can use an arbitrary platform such as Zenodo or Code Ocean. Please avoid using personal web pages.
- To be considered for the Best Paper Award, the authors must submit (at least) an Artifact Description appendix.
- If the paper used no computational artifacts, the authors need to respond “No” in the submission form.
- If the authors just choose to not provide computational artifacts, they need to respond “Wish not to provide” in the submission form.