Systems for Future Multi-Core Architectures
(SFMA'11)

EuroSys 2011 Workshop, Salzburg, Austria
Sunday 10th of April 2011

Introduction

The current trend towards multi-core computing is of significant importance to practitioners of systems-level software, such as operating systems, language runtimes and virtual machines. As the layer between application software and the underlying hardware, systems-level software must directly tackle the challenges of multi-core hardware (e.g., scalability, concurrency control and data-sharing costs), while providing appropriate abstractions to higher-level software. Future hardware is likely to increase the challenges encountered by systems software due to increasing system diversity, core heterogeneity, complex memory hierarchies, dynamic core failure, and non-cache coherent shared memory. However, the abundance of parallelism and potential for core specialization and inter-core message passing hardware also provide a number of new opportunities for system software. The current shift in hardware design provides an exciting opportunity to radically rethink the design and implementation of systems-level software.

The workshop on Systems for Future Multi-Core Architectures (SFMA'11) brings together researchers in the operating systems, language runtime and virtual machine communities to exchange ideas and experiences on the challenges and opportunities presented by future multi-core hardware. SFMA'11 will be co-located with EuroSys 2011 in Salzburg, Austria.

The proceedings for SFMA are now available to download, either as individual papers below, or as a single set of proceedings here.

Program

9:00Keynote
  Andrew Baumann - “Non-cache-coherent systems: The Barrelfish experience” (slides)
 
10:00Coffee
 
10:30Session 1: Parallel Runtime Systems and Operating Systems
  Session Chair: Tim Harris
  Scalable Lightweight Task Management Schemes for MIMD Processors (slides)
      Daniel Waddington, Chen Tian (Samsung R&D) and KC Sivaramakrishnan (Purdue University)
  OctoPOS: A Parallel Operating System for Invasive Computing (slides)
      Benjamin Oechslein, Jens Schedel, Jürgen Kleinöder (Friedrich-Alexander University Erlangen-Nuremberg),
    Lars Bauer, Jörg Henkel (Karlsruhe Institute of Technology), Daniel Lohmann and
    Wolfgang Schröder-Preikschat (Friedrich-Alexander University Erlangen-Nuremberg)
  Condensing the cloud: running CIEL on many-core
      Malte Schwarzkopf, Derek G. Murray and Steven Hand (University of Cambridge Computer Laboratory)
 
12:00Lunch
 
13:30Session 2: Concurrency
  Session Chair: Derek Murray
  Work in Progress: Evaluation of Parallel Design Patterns for Message Processing Systems on Embedded Multicore Systems (slides)
      Ronald Strebelow (Institute of Computer Science, University of Augsburg),
    Christian Prehofer (Fraunhofer Institute for Communication Systems ESK)
  Work in Progress: Shared-Nothing Transactional Memory
      Maurice Herlihy and Irina Calciu (Brown University)
  Integrating Dataflow Abstractions into Transactional Memory
      Vladimir Gajinov (Barcelona Supercomputing Center), Milos Milovanovic (Microsoft Development Center Serbia),
    Osman Unsal, Adrian Cristal, Eduard Ayguade and Mateo Valero (Barcelona Supercomputing Center)
  On the Nature of Progress (slides)
      Maurice Herlihy (Brown University), Nir Shavit (Tel Aviv University)
 
15:00Coffee
 
15:30Panel discussion
 
Topic:Exploiting Many-Core Architectures at the Systems Level
Moderator:   Joe Sventek (University of Glasgow)
Panelists:
Maurice Herlihy (Brown University)
Stephan Diestelhorst (AMD)
Paul Barham (Microsoft Silicon Valley)
Ross McIlroy (Microsoft Cambridge)
 
16:30Wrap-up
 
17:00Close

Paper submission:

Authors are invited to submit original and unpublished work that exposes a new problem, advocates a specific solution, or reports on actual experience. Papers should be submitted in the standard double column ACM SIG proceedings format, and are limited to 6 pages, including figures and tables. Papers that exceed this length will be rejected without consideration of their merit.

Final papers will be available to participants electronically at the meeting, but to facilitate resubmission to more formal venues, no archival proceedings will be published, and papers will not be sent to the ACM Digital Library. Authors will have the option of having their final paper accessible from the workshop website.

Paper submissions can be uploaded here.

If you have questions, please email: rmcilroy (at) microsoft (dot) com.

Important dates:

Submissions due: 10th February 2011 (11:59pm EST)
Notification: 1st March 2011
Camera-ready submission: 1st April 2011
Workshop: 10th April 2011

Organization:

Co-Chairs:

Ross McIlroy (Microsoft Research Cambridge)
Joe Sventek (University of Glasgow)
Tim Harris (Microsoft Research Cambridge)
Timothy Roscoe (ETH Zurich)

Program committee:

Ali-Reza Adl-Tabatabai (Intel)
Mats Brorsson (KTH Royal Institute of Technology, Sweden)
Juan Colmenares (University of California, Berkeley)
Stephan Diestelhorst (AMD)
Tim Harris (Microsoft Research Cambridge)
Hermann Härtig (TU Dresden)
Ross McIlroy (Microsoft Research Cambridge)
Derek Murray (University of Cambridge)
Timothy Roscoe (ETH Zürich)
Michael Scott (University of Rochester)
Jeremy Singer (University of Glasgow)
Joe Sventek (University of Glasgow)
Michael Swift (University of Wisconsin)
Ian Watson (University of Manchester)
Nickolai Zeldovich (MIT)