The presentations of the conference’s main track are now available on the programme page.

The registration page for EDCC is now open. We have also posted some additional information regarding the conference location, accommodation and Valencia in general.

Following a number of requests, we have extended the deadline for EDCC Fast Abstracts to 8 March (Monday next week).

Fast Abstracts provide an excellent opportunity to report on current work, to introduce new ideas or to formulate positions on controversial issues or open problems. Fast Abstracts typically foster lively discussion between participants and often help establish new contacts to interested peers.

Plese find further details on http://edcc.dependability.org/call-for/fast-abstracts

Prof. Jose Duato Marín from the Universidad Politécnica de Valencia will deliver the Second Keynote Speech at EDCC-2010 entitled:

“Addressing heterogeneity, failures and variability in high-performance NoCs”

Abstract: Although most research on NoCs has assumed the use of regular topologies like 2D meshes, some current trends in chip architecture, combined with expected technology limitations and usage models, will very likely oblige designers to consider less regular topologies to provide the best cost-performance trade-off. Moreover, the set of nodes interconnected by those NoCs will also be heterogeneous, including computational cores of different sizes and computing power, cache blocks and local stores, accelerators of different kinds, and memory controllers. The memory wall problem will likely be addressed by using 3D integration, which will increase heterogeneity significantly, due to the need for locating the hottest cores in the top layer. Manufacturing failures and variability will also introduce heterogeneity.

Therefore, in order to deliver the best cost-performance trade-off while minimizing resource and power consumption and providing the maximum flexibility, heterogeneity needs appropriate hardware support in the NoC. This talk motivates the need for efficiently supporting  heterogeneity, and sketches some results along this direction, describing power-efficient routing algorithms that provide support for multiple heterogeneous, possibly overlapping regions (e.g. virtual machines, coherence domains) in the presence of faulty components. The talk also shows how a hierarchical interconnect (on-chip, on-substrate) can significantly shorten design cost and time to market.

Speaker’s vita: Jose Duato received the MS and PhD degrees in electrical engineering from the Technical University of Valencia, Spain, in 1981 and 1985, respectively. Currently, Dr. Duato is Professor in the Department of Computer Engineering (DISCA) at the same university. He was also an adjunct professor in the Department of Computer and Information Science, The Ohio State University.

His current research interests include interconnection networks and multiprocessor architectures. Prof. Duato has published over 400 refereed papers. He proposed a powerful theory of deadlock-free adaptive routing for wormhole networks. Versions of this theory have been used in the design of the routing algorithms for the MIT Reliable Router, the Cray T3E supercomputer, the on-chip router of the Alpha 21364 microprocessor, and the IBM BlueGene/L supercomputer. Prof. Duato also developed RECN, the only truly scalable congestion management technique proposed to date, and a very efficient routing algorithm for fat trees that has been incorporated into Sun Microsystem’s 3456-port InfiniBand Magnum switch. Currently, Prof. Duato leads the Advanced Technology Group in the HyperTransport Consortium, which developed the High Node Count HyperTransport Specification 1.0 to extend the device addressing capabilities of HyperTransport in several orders of magnitude.

In 2009, Prof. Duato recently received the prestigious award “Julio Rey Pastor“, in the field of mathematics and information technologies and the communications, for his contributions of international importance in the field of the interconnection networks and for the transference of these results to the industry.

Presentation: pptx, pdf

[Post updated 17/5/2010]

The EDCC-2010 Programme and the list of accepted papers has now been announced. 25 papers were accepted, of which 19 regular papers, 4 practical experience reports, 1 industrial track papers, and 1 prototype/tool description paper.

First Workshop and Tool Session on DYnamic Aspects in DEpendability Models for Fault-Tolerant Systems (DYADEM-FTS 2010, Valencia, Spain, April 27 2010)

Rules

You may either submit:

• Novel tools: a two-page abstract describing your tool
• Established tools: a full-length submission already published elsewhere. In this case, your contribution will not appear in the DYADEM workshop proceedings.

All tools will be presented in a hands-on session during the workshop. For more details and the scope of the workshop, please refer to the workshop website.

Important dates

Submission deadline: 21 February 2010 23:59 GMT
Author notification: 1 March    2010
Camera ready due: 19 March    2010
DYADEM-FTS workshop date: 27 April    2010

New deadline for the Student Forum: 21 Feb 2010
New deadline for the DD4LCC workshop: 28 Jan 2010
New deadline for the CARS workshop: 27 Jan 2010

As in previous years, EDCC-2010 will host two highly popular events traditionally associated with the conference: the EDCC-2010 Fast Abstracts & the EDCC-2010 Student Forum. The Student Forum welcomes futuristic, sometimes “off the wall” ideas in the area of Dependable Computing from current Ph.D. and MSc students. Fast Abstracts are short presentations of work in progress or opinion pieces that aim to serve as a rapid and flexible mechanism to advance dialog and encourage interaction in the Dependability Community.

Submissions to he Student Forum are due by 21 February 2010, an to the Fast Abstracts by 28 February 2010.

EDCC-2010 will host three high-quality workshops on new and emerging areas in dependabilit on April 27, 2010, one day before EDCC’s main track. Check our workshops page to learn more about this.

New (15 Dec 2009): Papers accepted to the workshops will be published in the ACM Digital Library.

Brendan Murphy from Microsoft Research will deliver the Opening Keynote Speech at EDCC-2010:

“Developing accurate risk models requires mathematics, domain knowledge and common sense, although not necessarily in that order“

Abstract: It is very difficult to develop a risk model and have a paper describing that model being accepted in a journal or conference. But the real challenge is getting a set of experience engineers to accept and use your risk model. To achieve this you have to understand the limitations of the model and to do that you have to understand the limitations of the data you are analysing and also how the software is being developed. This talk looks at the development of a risk model for the Windows operating system, providing insights into our realization that mathematics does not necessarily solve all problem.

Speaker’s vita: Brendan is a Senior Researcher at Microsoft Research Centre in Cambridge UK. Brendan works in the Empirical Software Engineering and Measurement (ESM) activities in Microsoft focusing on software reliability, dependability, quality and process issues.

Presentation: pptx, pdf

[Post updated 17/5/2010]