Keynoters
KEYNOTERS - HPCS 2010
Tuesday Keynote I:
The Next Generation of High Performance Computing
William D. Gropp,
University of Illinois Urbana-Champaign, Illinois, USA
Tuesday Keynote II:
Practical Distribution of Random Streams for Stochastic High Performance Computing:
Application to Life Sciences
David Hill, Blaise Pascal University, France
Wednesday Keynote I:
On Cyber-Physical Systems Challenges and Research Opportunities
Taieb Znati,
University of Pittsburgh and National Science Foundation, USA
Wednesday Keynote II:
Building the European e-Infrastructure Ecosystem for Data Intensive Science
Robert Jones,
EGEE-I and EGEE-II Projects Director, CERN, Geneva - Switzerland
Thursday Keynote I:
Data-driven Modeling and Design of Networked Mobile Societies:
A Paradigm Shift for Future Social Networking
Ahmed Helmy,
University of Florida, Florida, USA
Thursday Keynote II:
Distributed Knowledge Discovery Services in Grids and Clouds
Domenico Talia,
Università della Calabria and ICAR-CNR, Italy
Tuesday Keynote I:
The Next Generation of High Performance
Computing
William D. Gropp,
University of Illinois Urbana-Champaign, Illinois, USA
ABSTRACT
Clusters revolutionized computing by making supercomputer capabilities widely
available. But one of the main drivers of that revolution, the rapid doubling of
processor clock rates, ran out of steam several years ago. To maintain (or even
increase) the historic rate of improvement in computing power, processor designs are
rapidly increasing parallelism at all levels, including more functional units, more cores,
and ways to share resources among threads. Heterogeneous designs that use more
specialized processors such as GPGPUs are becoming common. The scale of high-end
systems is also getting larger, with 1000-core systems becoming commonplace and systems
with over 300,000 cores becoming operational in 2011. However, the software and
algorithms for these systems are still basically the same as when the cluster revolution
began. Drawing on experiences with the sustained PetaFLOPS system, called Blue
Waters, to be installed at Illinois in 2011, and with exploratory work into Exascale
system designs, this talk will discuss some of the challenges facing the high performance
and cluster community as scalability becomes increasingly important and reviews some of
the developments in algorithms, programming models, and software frameworks that must
complement the evolution of high performance computing hardware.
Tuesday Keynote II
:
Practical Distribution of Random Streams for
Stochastic High Performance Computing: Application to Life Sciences
David Hill
, Blaise Pascal University, France
ABSTRACT
The purpose of this talk is to present the best practical approaches used to distribute
random streams in the context of Stochastic High Performance Computing. The first
part will give a short presentation dealing with the differences between various
paradigms: Grid Computing, Internet Computing and Cloud Computing, Peer-To-Peer, Cluster
Computing, Meta Computing, Parasitic computing and Hybrid Computing. Then, in a
second part, the focus will be given to the distribution of stochastic applications and to
the distribution of simulation experiments (whether stochastic or not). The best
techniques in use to distribute pseudo random numbers will be presented as well as future
research directions. A highlight will be given to optimization techniques and real
applications on ecological modelling and medical projects will be presented. The
case of successful large-scale simulations and data challenges on the EGEE Grid will also
be exposed.
Wednesday Keynote I
:
On Cyber-Physical Systems Challenges and
Research Opportunities
Taieb Znati
, University of Pittsburgh and National Science Foundation, USA
ABSTRACT
Unprecedented advances in wireless and mobile technologies, coupled with the proliferation
of social networks and applications, is paving the way for a new era of cyber-physical
systems that will revolutionize the way humans interact with their physical environment.
Cyber-physical systems exhibit deeply integrated computational and physical capabilities,
interacting with humans through diverse modalities. The ability to interact with and
expand capabilities of the physical world through computational means is the key
technological multiplier. Realizing the vision of cyber-physical systems, however, brings
about unprecedented challenges, technical and non-technical, underscoring the need for
radical thinking and new insights into how we design and deploy distributed cyber-physical
systems, on which our lives and critical sectors of our society can depend. These systems
are likely to exhibit complex dynamics at different spatial and temporal scales, and
various levels of control. They need to be predictive, reactive to conditions and external
events, and receptive to coordination and negotiation. They also need to be fault tolerant
and recoverable, satisfying potentially very high availability and timeliness
requirements. As critical as these physical and cyber infrastructures are to our lives and
diverse sectors of our society, we have little rigorous knowledge for understanding their
structure and dynamics. The talk will discuss challenges and future research directions in
how to effectively design robust and secure large-scale complex systems, so that we can
engineer them to have predictable behaviors.
Wednesday Keynote II
:
Building the European e-Infrastructure
Ecosystem for Data Intensive Science
Robert Jones
, EGEE-I and EGEE-II Projects Director, CERN, Geneva - Switzerland
ABSTRACT
The importance of research data for modern science is growing daily, and new initiatives
are been required to cope with the resulting "data deluge". The emergence of big and
complex-data science is here to stay. It will open completely new ways to dig
knowledge out of the huge amounts of information that are becoming available across a
range of research from astronomy to archaeology, and physics to epidemiology.
Incorporating e-Science digital repositories and their holdings into an open information
ecosystem will help support new scientific methods and paradigms, improving both the
efficiency of the scientific process and its impact.
Existing Pan-European computing infrastructures, including high-speed networking (GEANT),
high-capacity grid systems (EGEE/EGI) and specialised high-performance computing centres
(DEISA/PRACE) serve a significant proportion of Europe's research community.
By considering these developments, this talk will suggest some opportunities and
challenges for how e-infrastructures can evolve in the future to address the challenges
facing Europe's research communities.
Thursday Keynote I
:
Data-driven Modeling and Design of Networked
Mobile Societies: A Paradigm Shift for Future Social Networking
Ahmed Helmy
, University of Florida, Florida, USA
ABSTRACT
The future of social networking is in the mobile world. Future network services are
expected to center around human activity and behavior. Wireless networks (including ad
hoc, sensor networks and DTNs) are expected to grow significantly and accommodate higher
levels of mobility and interaction. In such a highly dynamic environment, networks need to
adapt efficiently (performance-wise) and gracefully (correctness and functionality-wise)
to growth and dynamics in many dimensions, including behavioral and mobility patterns,
on-line activity and load. Understanding and realistically modeling this multi-dimensional
space is essential to the design and evaluation of efficient protocols and services of the
future Internet.
This level of understanding to drive the modeling and protocol design shall be developed
using data-driven paradigm. The design philosophy for the proposed paradigm is unique in
that it begins by intensive analysis of measurements from the target contexts, which then
drive the modeling, protocol and service design through a systematic framework, called
TRACE. Components of TRACE include: 1. Tracing and monitoring of behavior, 2. Representing
and Analyzing the data, 3. Characterizing behavioral profiles using data mining and
clustering techniques, and finally 4. Employing the understanding and insight attained
into developing realistic models of mobile user behavior, and designing efficient
protocols and services for future mobile societies.
Tracing at a large scale represents the next frontier for sensor networks (sensing the
human society). Our latest progress in that field (MobiLib) shall be presented, along with
data mining and machine learning tools to meaningfully analyze the data. Several
challenges will be presented and novel use of clustering algorithms will be provided.
Major contributions to modeling of human mobility (the time variant community model,
TVC) will also be discussed.
Finally, insights developed through analysis, mining and modeling will be utilized to
introduce and design a novel communication paradigm, called profile-cast, to support new
classes of service for interest-aware routing and dissemination of information, queries
and resource discovery, trust and participatory sensing (crowd sourcing) in future mobile
networks. Unlike conventional - unicast, multicast or directory based - paradigms, the
proposed paradigm infers user interest using implicit behavioral profiling via
self-monitoring and mining techniques. In order to capture interest, a spatio-temporal
representation is introduced to capture users behavioral-space. Users can then identify
similarity of interest based on their position in such space.
The proposed profile-cast paradigm will act as enabler to new classes of service, ranging
from mobile social networking, and navigation of mobile societies and spaces, to emergency
alerts and disaster relief using infrastructure-less networks, among others.
Thursday Keynote II
:
Distributed Knowledge Discovery Services in
Grids and Clouds
Domenico Talia
, Università della Calabria and ICAR-CNR, Italy
ABSTRACT
Grid and Cloud computing systems were originally designed for dealing with problems
involving compute-intensive tasks. Today, however, Grids and Clouds enlarged their horizon
as they are going to run both scientific and business applications using large data
sources for scientists, professionals and end users. To face those new challenges, such
distributed computing environments must support adaptive knowledge discovery and data
mining applications by offering resources, services, and decentralized data analysis
methods. In particular, according to the service oriented architecture (SOA) model, data
mining tasks and knowledge discovery processes can be delivered as services in Grid and
Cloud computing infrastructures.
Through a service-based approach we can define integrated services for supporting
distributed scientific data analysis tasks in Grids and Clouds. Those services can address
all the tasks that must be considered in knowledge discovery processes from data selection
and transport, to data analysis, knowledge model representation and visualization. We are
working along this direction by providing service-oriented architectures and
services for distributed knowledge discovery.
This collection of data mining services composes an Open Service Framework for Distributed
Knowledge Discovery. This framework allows developers to design distributed KDD processes
as a composition of services that are available over HPC computers and large scale
distributed infrastructures (e.g., from a single Cloud to Interclouds).
Here we describe a strategy based on the use of services for the design of open
distributed knowledge discovery services and outline how Grid and Cloud application
frameworks can be developed as a collection of services.