The 5th International Workshop on Adaptive Self-tuning Computing Systems

(co-located with HiPEAC 2015)

21 January 2015, Amsterdam, The Netherlands

Best paper award is sponsored by

Continuing effort on reproducibility is sponsored by




News:

Overview

Computing systems are rapidly evolving into heterogeneous machines featuring many processor cores. This leads to a tremendous complexity with an unprecedented number of available design and optimization choices for architectures, applications, compilers and run-time systems. Using outdated, non-adaptive technology results in an enormous waste of expensive computing resources and energy, while slowing down time to market.

The 5th International Workshop on Adaptive Self-tuning Computing Systems is an interdisciplinary forum for researchers, practitioners, developers and application writers to discuss ideas, experience, methodology, applications, compilers, hardware, practical techniques and tools that can improve or change existing and future computing systems using self-tuning technology. Such systems should be able to automatically adjust their behavior at all levels (hardware and software) and improve various characteristics including execution time, energy usage, reliability, size and costs based on empirical, dynamic, iterative, auto-tuning, statistical, collective, bio-inspired, machine learning and alternative techniques while fully utilizing available resources.

Along with the traditional publication model, we propose a new open publication model where community pre-reviews articles, shared artifacts and experimental results [ docwiki ]!


Submission guidelines

We invite papers in three categories:
  • T1: Unpublished Papers presenting novel unpublished ideas (did not appear in any proceedings). Depending on the maturity of the work, these papers should be between 2 and 6 pages long (excluding bibliography). In addition to mature ideas, preliminary and exploratory work is also welcome in this category, including wild & crazy ideas.
  • T2: Already Published Work (must be your own) presenting ideas relevant to the ADAPT workshop. These papers should be up to 2 pages long (excluding bibliography) summarising the contribution of the original work (including link to it).
  • T3: Software/Hardware Demonstration (must be your own) presenting publicly available and adaptive software and hardware. Description should be up to 6 pages long summarizing the tool, demonstration, and relevance to the workshop.
Submissions should be in PDF formatted with double column/single-spacing, using 10pt fonts and printable on US letter or A4 sized paper. Accepted papers will be published online on the conference website and it will not prevent later publication of extended papers. Depending on the author wish, we may consider arranging ACM digital library proceedings or a special journal issue.

Paper submission website: https://easychair.org/conferences/?conf=adapt2015

For any questions/problems related to the submission please send an email here: christophe.dubach (_@_) ed.ac.uk and grigori.fursin (_@_) cTuning.org.

Optional Community-Driven Evaluation

Although not mandatory, we encourage authors to opt-in for our new experimental publication model where papers, artifacts and experimental results are pre-validated by the community. If you wish to participate in this experiment, simply submit your paper as usual on the submission website and send an email to the PC chairs indicating you wish to take part. We will give you instruction about how to upload your paper to arXiv and we will open a discussion thread on the reddit social platform to allow anyone to evaluate and comment on it (the authors will be allowed to respond to all the comments at any time). One week before the notification deadline, we will close the discussion thread and the members of the program chair will reach their decisions based on their own evaluation but also from the comments of others (including the authors) found on the discussion thread. We hope that this will help us make the reviewing process more fair, transparent and interactive while picking up papers that can raise interesting discussions at the workshop as described in our vision paper available in ACM DL and arXiv.

Optional Artifact Evaluation

To continue our public effort started in 2007 to enable collaborative, systematic and reproducible research and experimentation on auto-tuning combined with machine learning, crowdsourcing and run-time adaptation, we particularly welcome papers where all related research material including code, data and experimental results is shared and validated by the community. Therefore, we encourage authors to either provide a paragraph in their submitted papers describing how to validate their experimental results or to submit artifacts for validation after their paper is accepted. See our reproducibility wiki for more details.

Workshop organisers / program chairs

Program committee

Artifact evaluation

Artifact evaluation is chaired by Grigori Fursin and Tom Crick. Authors of accepted papers can voluntarily submit their artifacts for evaluation. The validation procedure is the same as at previous ADAPT workshops as well CGO'15 and PPoPP'15. If you want to receive our news, follow us on our mailing list, Twitter, Facebook, or LinkedIn!

IMPORTANT DATES

  • Paper submission: 24 October 2014 (no extension)
  • Notification: 19 November 2014
  • Final version: 3 December 2014
  • Early fees registration: 23 December 2014
  • Workshop: 21 January 2015

  • If you have questions, comments or suggestions, do not hesitate to contact organizers!