Keynote Speakers
We are very pleased to have acquired the services of an excellent selection of keynote speakers for the ICCS'12. Note that the list is incomplete at the moment and other speakers will be added over time.
Matjaz Mulej |
Brief Bio:
Matjaz Mulej was born in 1941 in Slovenia, he is a Ph.D. in economic sciences (specialised in systems theory) and in management science (in the field of innovation management), professor emeritus of systems and innovation theories. He works at the Faculty of Economics and Business in Maribor and the author of the dialectical system theory and the model of innovative business theory (for economy in transition). He has published more than 60 books and publications in more than 40 countries and worked as a visiting professor and researcher in 6 different countries for 15 semesters, including Cornell University, USA. He is a member of three international academies of sciences and arts (Salzburg, Paris, and Vienna) and the president of the International Academy of Systems and Cybernetic Sciences. Until April 2010, he was the president of IFSR (with 37 member associations and members from all over the world). He appears under number 8082 in the Researcher Register with more than 1600 publications being cited around 120 times and publishing more than 50 articles with co-authors in foreign first-class magazines. He was a chancellor and a vice-chancellor of the University of Maribor from 1997 to 2001. He worked as an innovation consultant almost 500 times in Slovenia as well as the USA, Russia, Angola, Indonesia, and Poland. In the International Encyclopaedia of Systems and Cybernetics, 2nd edition (Charles François, 2004. München: Saur), he and his dialectical system theory is named nine times and his name is frequently searched online as well. Lately, he has been encouraging the development of social responsibility as a possible solution to human crisis, as he regards it as an informal way to the sufficient and necessary integrity, in other words to innovate ethics with system behaviour. This is one of the reasons why he works as the head of the Expert Committee of the Institute for Social Responsibility Development in Maribor.
Abstract:
Ludwig von Bertalanffy (1978, p. VII) explicitly stated that he had created his General Systems Theory against over-specialization, i.e. to support interdisciplinary creative cooperation as the best way toward the necessary holism of approach and wholeness of outcomes of human activity. But he did no support this intention methodologically a lot.
The now-a-days usual narrow over-specialization provides for no/poor humans capacity of interdisciplinary creative cooperation. Earlier it was much less dangerous than today, as the current socio-economic crisis shows. Narrow specialization is still necessary, but equally so is the other specialists’ capacity: cooperation helps humans prevent oversights and resulting failures, because it enables more holistic thinking/behavior.
But there are very few humans around the world, who can/may teach holistic thinking. The role of the narrow specializations which is unavoidable is so strong that people hardly see that holistic thinking – enabled by interdisciplinary creative cooperation – makes specialization of any profession much more beneficial than any operation of the specialization alone. Nobody, whatever their profession, can live well without co-operation with people of other professions. De Bono’s ‘6 Thinking Hats’ support it, so does Mulej’s Dialectical Systems Theory (DST) from the same period of time.
Some fifty years after the authors of Systems Theory had succeeded in making their theory known, and politicians of the world succeeded in using it (informally) by making the United Nations Organization the most holistic political organization of humankind. The European Union (EU) found it necessary to explicitly link ‘systemic’ views with innovation. The EU, after reminding readers of its previous documents enhancing innovation, states on page 6:
‘The Action Plan [First Action Plan for Innovation in Europe, 1996, based on Green Paper on Innovation, 1995] was firmly based on the ‘systemic’ view, in which innovation is seen as arising from complex interactions between many individuals, organizations and environmental factors, rather than as a linear trajectory from new knowledge to new product. Support for this view has deepened in recent years (EU, 2000)’.
Such a move to support and even require systemic thinking is taking place currently again under the label of social responsibility: United Nations Organization works on UN Global Compact for a decade, International Standards Organization launched ISO 26000 (ISO, 2010), and European Union advises its member states and big enterprises to use ISO 26000 as a way out from the current socio-economic crisis (EU, 2011).
Thus, the concepts of ‘interdependence’ and ‘holistic approach’, i.e. systemic behavior, are found crucial on the world-top level.
If this has to be stated explicitly in such documents, these questions arise:
- Are we humans capable of the interdisciplinary co-operation that we need almost every moment?
- What is the theoretical basis for those, who are not capable of it, to learn?
Very few humans are by their nature and education capable of interdisciplinary co-operation, because specialists teach specialists to be specialists, including being proud of their specialization. This teaching is reasonable, but it is not enough: it may cause one to hide from reality behind the walls of one’s specialization and lack respect for other specializations and their need for each other - as well as restricting their capacity to solve real problems by interdisciplinary creative co-operation much better than by separation (Ackoff, 2001, 2003; Gigch, 2003; Mulej, 1974, 1979; Mulej et al, 2012).
The theoretical basis to learn the skills of the interdisciplinary co-operation stems from the original authors of the Systems Theory and Cybernetics: Bertalanffy and Wiener. But many humans forget that the fathers of the Systems Theory and Cybernetics have created their answers to the burning problems of their and our time through their interdisciplinary approach. This is where our Dialectical Systems Theory (DST) (Mulej, 1974; 1979; 1992; 2000; 2012) originating nearly 4 decades ago, allows us to fill the gap. The well-intended and well-applied versions of systems theory, which describe a part of reality inside the viewpoint of one or another traditional, specialized, scientific discipline, are beneficial, but they do not match the well stated EU’s and others’ definition of ‘systems view’. Thus, they help people solve other problems, but not that of the holism of thinking, decision-making, and action, as a precondition of survival of humankind and the planet on which we live, and/or of success in any human action. (Geyer et al, 2003). Interdependence of different professions is left aside. We will try to demonstrate here that holism of thinking, decision-making, and action is very necessary and has been the point of Ludwig von Bertalanffy (1950, 1978) and his co-establishers of the General Systems Theory and related movement. It is even more needed today, and has been very much forgotten about, but it can be practiced well, if people use what we call DST (See also: François, 2004; he calls DST peculiar, for this reason, obviously). ------------------------------------------------------------------------------------------------------------------------------------------------------------
Pierre Bricage |
Brief Bio:
Pierre Bricage was born in 1947 in Paris, France. As alumnus of the ENS de Saint Cloud (Lyon), France,
he passed the aggregation of biology. As a biologist, he has edited or published more than 250 pedagogic
or scientific works in more than 20 countries (http://web.univ-pau.fr/~bricage/). During 8 years at the
University of Dakar (Sénégal), the biological rhythms (experimental study and modelling) of
biochemical, ecological, physiological & genetical markers of plant enzymes & pigments were about his
teaching & research first interests in Biology and Ecology (sustainable management of natural resources
and environmental education). He has taught Systems Theories & Micro-Informatics (Data Bases
modelling, programming, simulating) applied to Health and Social Sciences (Societal Engineering and
Man's Societal and Environmental Responsibility). Since 2000, he is pointing back to Fundamentals in
Biology & Systemics Practical Applications (http://hal.archives-ouvertes.fr/hal-00130218/fr) with all
OnLine “Creative Commons” works. He has been developing new Methods & Key Words in biosystemics
(endophysiotope and ecoexotope, “the gauge invariance of life“, phylotagmotaphology
http://hal.archives-ouvertes.fr/hal-00423730/fr), through the new predictive Paradigm of ARMSADA
the “Associations for the Reciprocal and Mutual Sharing of Advantages and DisAdvantages”
http://www.armsada.eu. Now retired, he is Associate Secretary General of the French Association of
Systemics and Cybernetics AFSCET (http://www.afscet.asso.fr), Deputy Secretary General of the
European Union for Systemics UES-EUS (http://ues-eus.eu/), Member of the Directorate of the World
Organisation of Systems and Cybernetics WOSC (http://www.wosc.co/) and Secretary General of the
International Academy for Systems and Cybernetic Sciences IASCYS (http://www.iascys.org).
Time Management by Living Systems. A General System Theory of the Time Modularity of the Living Systems: Modeling How Zeitgebers Interactions, Clocks, and Rhythms Design Conics Running Calendars.
Abstract:
The holistic approach of the description of a living system is based on the concept that the system participates not only to its evolution but also to that of all the sub-systems it contains and all the systems in which it is a sub-system (Fig. 1). No system can be defined as an independent, organizationally closed space but rather as an independent, autonomous closed “time and timer“. Its time modularity is revealed by rhythms at different scales. The clocks of the system result from the recall of both external and internal time information. The watches, clocks and calendars, which are the actors of the endogenous rhythms, are built as a result of the memorization of past responses of interaction (phase shift delays, entrainments, breakages) between its endophysiotope and ecoexotope. The system needs to be both a clock-maker, a watch-maker and a wake-maker. So modeling has to take into account “simultaneously“ the concepts of [1] -temporal window: "Before the time, that is not the time. After the time, this no longer is the time.", -time latency and time shifting: "It is necessary to give some time to the time.", -compartmentation of the time: "There is a time for each event. And each event is located into its time place.", -nonlinearity and non-summation: "The temporal Whole is both more and less than the sum of its Parts." and -interactions between time and space: “The arrow of the time structures the clocks of the living systems. But, the living systems structure back the arrow of their time.“ [2]. Calendars are tools for forecasting gates (allowed time zones of a “cycle“ through which an event may emerge) and fixing a time horizon (a fixed point in the future at which a processes will be evaluated or be assumed to end or start). The time knowledge needs the design by the living system of a range of skills and tools used to manage time when accomplishing specific tasks to survive. It appears that timelines are designed as ellipses, projected on a plan or a Moebius strip, labeled with dates alongside and events labeled on the points where they would have happened. The emergence of a new blueprint runs through the juxtaposition and embedment of previous systems. The new time Whole is both more and less than the sum of its Parts (Table 1), it merges through the simultaneous metamorphoses of the Parts into the Whole. But each subsystem maintains its space-time identity into the Whole of which it is a partner. The partial autonomy of each partner is allowed through the maintenance of individual or collective, spatial and temporal boundaries. These interfaces structure the spatial and temporal integration of the parts into the endophysiotope (ENDO) of their whole, and of the whole into the ecoexotope (EXO) of its survival. The transition from one level of organization to an adjacent and superior one is the result of the building of a new, spatial and temporal, network. In this new orderly spaced-timed system all braces are allowed and each partner owns a special place, both through the time and into the space. The integration of the parts, and simultaneously of the whole, merges through the building of Associations for the Reciprocal and Mutual Sharing of Advantages and DisAdvantages (http://armsada.eu). This allows, in a no-change ecoexotope (Fig. 6), the maintenance of both the requisite variety of the partners and the unity of the whole [3]. If changes of the endophysiotope or ecoexotope, this is the only way to make a new networking mode of organization and integration. These associations merge through the interactive fitness between the capacity "to welcome" of the ecoexotope and the capacity "to be welcomed" of the endophysiotope of each parcener. Failures of medical treatments and pathological processes can be explained in terms of “irregularities” or breakages of the time architecture. Inside a system, the time can stop, move with different speeds, jump from one step to another, but it will never turn back. When a step is in the past you need to turn forward a complete cycle to reach it again (anthocyanin adaptive synthesis, asthma crises, rhythms of peroxidase capacities, glycaemia regulation, awakenings times, forest equilibrium).
------------------------------------------------------------------------------------------------------------------------------------------------------------
Tarek El-Ghazawi, |
Brief Bio:
Tarek El-Ghazawi is a Professor in the Department of Electrical and Computer Engineering at The George Washington University, where he leads the university-wide Strategic Academic Program in High-Performance Computing. El-Ghazawi’s research interests include high-performance computing, computer architectures, reconfigurable, embedded computing and computer vision. He is one of the principal co-authors of the UPC parallel programming language and the first author of the UPC book from John Wiley and Sons. He has received his Ph.D. degree in Electrical and Computer Engineering from New Mexico State University in 1988. El-Ghazawi has published well over 200 refereed research publications in this area. Dr. El-Ghazawi has served in many editorial roles including an Associate Editor for the IEEE Transactions on Computers. He has chaired and co-chaired many international conferences and symposia including, The 10th IEEE International Conference on Scalable Computing and Communications (ScalCom-10), 2010 and the 24th IEEE Conference on Applications-Specific Systems, Architectures and Processors (IEEE ASAP2013). Dr. El-Ghazawi’s research has been frequently supported by Federal agencies and industry including DARPA/DoD, NSF, DoE/LBNL, AFRL, NASA, IBM, HP, Intel, AMD, SGI, and Microsoft. Professor El-Ghazawi is a Fellow of the IEEE and a Research Faculty Fellow of the IBM Center for Advanced Studies, Toronto. He is a member of the Phi Kappa Phi national honor society and an elected member of the IFIP WG10.3. El-Ghazawi was a recipient of the 2012 Alexander Schwarzkopf Prize for Technical Innovation.
For more information: http://www.seas.gwu.edu/~tarek/
Mobile Cloud Computing: Opportunities and Challenges
Abstract:
The emergence and rapid spread of interest and use of cloud computing as an accessible and expandable, as needed, computing facility on the go, has a very deep affinity to the proliferation of intelligent mobile devices including smart phones and tablets. Together, these technologies have the potential of not leaving anybody behind when it comes to computing applications whether small and personal or large and organizational, and regardless of geographic boundaries and economical conditions. However, many technical challenges still exist that are still delaying the realization of this dream with the responsiveness and quality needed from the user perspective. In this talk we examine user requirements for access to clouds through hand held devices. In the light of those requirements we characterize some of the needed research developments particularly in the areas of device architecture and exchange protocols. In some of those areas we also present initial findings and we will highlight potential future directions. .
------------------------------------------------------------------------------------------------------------------------------------------------------------
Luis Martínez López |
Brief Bio:
Luis Martínez was born in 1970. He received the M.Sc.and Ph. D. degrees in Computer Sciences, both from the University of Granada, Spain, in 1993 and 1999, respectively
Currently, he is Full Professor of Computer Science Department at the University of Jaén. His current research interests are linguistic preference modelling, decision making, fuzzy logic based systems, computer aided learning, sensory evaluation, recommender systems and electronic commerce. He co-edited nine journal special issues on fuzzy preference modelling, soft computing, linguistic decision making and fuzzy sets theory and published more than 40 papers in journals indexed by the SCI as well as some book chapters and more than 100 contributions in International and National Conferences related to his areas.
It is remarkable that he has been main researcher in 11 R&D projects.
He is member of the European Society for Fuzzy Logic and Technology, IEEE, Editor in Chief of the International Journal of Computer Intelligence Systems, Manager Editor of the journal Soft Computing and member of the editorial board of the Journal of Universal Computer Science.
He has the following international rankings:
- ISI Essential Science Indicators - Most Cited Scientists in Engineering
November 1, 2011 to cover a 10-year + 8-month period, January 1, 2001-August 31, 2011.
(Top 1%, A total of 7031 in the 1% of the most cited authors list).
L. Martínez Position: 1409. Number of citations: 743 High Cited Papers: 6
- h Index (ISI Web of Science): 16
For more information: http://www.di.ujaen.es/~martin
Qualitative information in Complex Decision Making
Abstract:
Decision Making is a core area in different fields in real world. In such fields, decision problems have been successfully solved by different existing decision models. However new needs in decision problems demands more realistic decision models able to deal with vagueness and imperfect information that takes part of the framework of complex real life decision problems.
The role of fuzzy sets have already played a key role in decision analysis solving processes since the seminal paper published by Bellman and Zadeh in 1970, but still they are an important basis to provide new tools, methods and methodologies in order to manage the inherent vagueness of preferences and imperfect information that provide decision makers in many situations in which they describe their knowledge by means of terms belonging to the natural language.
Many proposals based on different fuzzy techniques have been presented to deal with qualitative information in decision making providing valuable results in diverse decision situations, but recent research has introduced new directions, trends and models that can provide very promising results in the coming future in complex decision problems.
This talk will highlight current achievements and point out new trends and recent proposals that might be very useful to deal with qualitative information in those decision making problems that demands more realistic decision models.
---------------------------------------------------------------------------------------------------------------------------------------------------------
Antonio Caselles Moncho |
Brief Bio:
Antonio Caselles-Moncho was born in 1945 in Valencia (Spain). Agricultural Engineer in 1970. Master in Hortofruticultura I.A.M.Z. (OECD grant),
Bussiness Administration (Universidad Politécnica de Madrid), Operations Research (Universitat de València). Doctor Engineer in 1978.
Teching Mathematics in the "Universitat de València" since 1974. Teaching Operations Research an Systems since 1977. His research interests,
directed thesis, publications and professional works are related with Thistems Thinking, Complex Systems, Mathmatical Modelling and Simulation,
Automatic Programming and Data Mining. Director of the Operations Research School of the University of Valencia between 1992 and 1997. Foundation member
of the Spanish Society of General Systems (Sociedad Española de Sistemas Generales: SESGE). Vice-president of the SESGE since1994. Editor in chief of the
Revista Internacional de Sistemas (ISSN 0214-6533) since 1999.
For more information: http://www.uv.es/~caselles
Globalization and Sustainability: can the present globalization process be reoriented?
Abstract:
What we understand by globalization?
What we understand by sustainability?
How is the present process of globalization running?
- What is happening? Its advantages. Its disadvantages.<.li>
- How it is evaluated. Its repercussions on well-being and sustainability.
What can be made to avoid or to palliate its disadvantages without resigning to its advantages?
- Search for the causes of such disadvantages.
- Search for possible actions and their viability to try to eliminate or to palliate the causes of such disadvantages.
- To improve the ethics of the human beings.
- Legislation, control (monitoring of the real situation: data, statistics, and their diffusion) and rewards or punishments.
- Mathematical/logical modelization and simulation of the globalization process including its own assessment. Advantages and disadvantages of the method. Alternative methods.
- Review of the basic characteristics of the existing models and methods applicable to optimize the globalization process.
- Determination of the characteristics of a model or method that is specific for this problem (all modelization is perspectivistic): quantitative and qualitative, stochastic, several approaches, etc.
- Presentation of the initial works in this line promoted by Lorenzo Ferrer and their present situation.
- Quantitative model.
- Qualitative model.
-------------------------------------------------------------------------------------------------------------------------------------------------------------
Francisco Parra Luna |
Brief Bio:
Francisco Parra Luna was born in 1937. He studied Economics and Business Administration and
worked as Controller and Administrative Director en La Cruz del Campo, S.A., in Seville, and Cerveceras
Asociadads, in Barcelona. He has doctorates in Sociology and Political Science from the Universities of
Lausanne and Geneve (Switzerland), and the Universidad Complutense of Madrid. He is Catedratico de
Sociologia in this University. He created the "Working Group on Systems Theory" in the International
Sociological Association" and participated as Chairman of several international meetings. He is a member
of several academic associations and belongs of the editorial board of several journals on systems theory
and sociology. He has published about fifty scientific articles in professional journals and several books,
among them, Towards Comparing National Social Performances (University of Lausanne, 1974);
Balance Social y Progreso Empresaria (Edit. Cirde, Madrid, 1980); Sistema Sociopolitico y Seguridad
Social en España (Ed. Index, Madrid, 1980); Los Emigrante en españoles en Francia (Instituto Esp. de
Emigración, Madrid, 1981); Sociología Industrial y de la Empresa (written with J. A. Garmendia and M.
Navarro, Edit. Aguilar, Madrid, 1987); Politica de Empleo y Bienestar Social (Ed. Eudema, Madrid,
1988); El Balance Social de la Empresa (Ed. Deusto, Bilbao, 1989); Sociologia de la Empresa y de los
Recursos Humanos (Taurus, Madrid, 1993); La Empresa contra si misma (Ed. Deusto, Bilbao, 1993);
Sustainable Development (written with J. L. Elohim and E. Stuhler, Rainer Hampp Verlag, Munchen,
2000); and The Performance of Social Systems (Kluwer Academic/ Plenum Publishers, New York, 2000).
For more information: http://www.parraluna.es/
The Economic Crisis in Spain as a Desequilibrium Between Values: The Axiological-SETCU Model as a Solution
Abstract:
The world economic situation has significantly changed, not only because of the crisis, but mainly due to the fact that some classical economic principles, before centered around the production and distribution of real goods, are now more centered around an over-circulation of complex financial documents (MBS,CDO, CDS…….) which are not productive at all, except for a minority of experts and institutions, like some “shadow bank systems” which are not sufficiently regulated. In this context, our politicians and technicians are not taking into account the very high level of unemployment in Spain and its economic and social consequences. The three main axiological disequilibria (“economic competitive needs”, “unemployment” and “financial resources available”), much greater in Spain than in other developed countries, are clearly not interrelated in their economic analyses, whereas the systemic integration of these three disequilibria could solve a substantial part of the unemployment problem. The main reasons for this omission seems to be: first, that technicians are not ready to make the necessary efforts to adopt new theoretical and non-obsolete approaches to cultural and economic problems (for instance, the axiological and systemic approaches); and, second, the unethical behavior of politicians who usually think more of their personal and electoral interests than of the population’s needs. The attitude of these two main stakeholders could be at the heart of the grave Spanish unemployment problem.
----------------------------------------------------------------------------------------------------------------------------------------------------------
Mohamed Bakhouya |
Brief Bio:
Bakhouya is a senior research scientist
at Aalto University (Finland). He received the Ph.D. (2005) degree in computer
science and computer engineering from the University of Technology of Belfort Montbéliard
(UTBM), France. Between 2006 and 2008, he was a research scientist at George
Washington University (USA). He was involved in sponsored projects with CHREC, a
national NSF center of High-performance and REConfigurable Computing. Between
2008 and 2011, he was a research scientist at UTBM as co-PI of FP7 European
projects such as ASSET (Advanced Safety and Driver Support in Efficient Road
Transport, FP7-SST, 2008-2011), and TELEFOT (Field Operational Tests of
Aftermarket and Nomadic Devises in Vehicles, FP7-ICT, 2008-2012). He is also involved
in the EU EACEA Erasmus Mundus project TARGET (Transfer of Appropriate
Requirements for Global Education and Technology, 2011-2014), EU EACEA-LLP-Leonardo da Vinci VAPVoS (Virtual
Academy Platform for Vocational Schools, 2011-2013) and EU FP7 NoE, HYCON2 (Highly-complex
and networked control systems HYCON 2, 2010-2014, associate member, level 2), coordinated
by SUPELEC, Paris, France. His research interests include various aspects on
the design, validation, implementation, performance evaluation and analysis of
distributed systems, architectures, protocols and services. He serves as a
guest editor of a number of international journals, such as the ACM
Transactions on Autonomous and Adaptive Systems and ACM/Springer Mobile Networks
and Applications Journal.
For more information: http://mbakhouya.free.fr/
Collective Adaptive Systems: Approaches and challenges
Abstract:
The emerging paradigm of collective adaptive systems applies to
a variety of organizations, communities, domains and systems that include
many heterogeneous entities (e.g., human, computers, agents, devices,
services, sensors) cooperating in complex manner to achieve a common goal,
or even competing for conflicting objectives. Entities are operating at
different temporal, spatial scales and social scope and can dynamically
join or leave the collective at any time. Decision-making is also
distributed and influenced by the emergence of unexpected phenomena due to
the behavior of nodes/units of the collective. A typical example is
transportation system networks where decision-making is distributed and
highly time-location dependent. Units are heterogeneous and could join or
leave the collective at any time. Interactions between the units may lead
to the emergence of spontaneous phenomena. Unexpected behaviors can also
emerge from peered interactions, but information and context-awareness can
be partial, widely distributed and needs to be updated frequently.
A grand challenge is being able to design these complex systems and
incorporates mechanisms that allow them to learn, adapt, and self-organize
by redesigning their structure and their behavior to best suit human needs.
These mechanisms should handle dynamic environments in which resource
availability, user behavior and context change continuously over time. In
this talk, existing work dealing with designing and operating principles
for developing this type of complex systems will be presented. Emphasis
will be put on adaptive approaches mainly developed to address the
constantly changing of system’s environment. Some scenarios developed under
the EU project ASSET (http://www.project-asset.com/) will be briefly
presented.
Pierre Manneback |
Brief Bio:
Pierre Manneback obtained his Ph.D. in Applied Mathematics in 1985 at the University of Namur, Belgium. He is working since 1990 at University of Mons (UMONS), Faculty of Engineering, where he is now Professor in Computer Science. His current research is focused on parallel and distributed computing, mainly on heterogeneous systems with GPU. He is member of several scientific committees of international conferences in distributed computing (IPDPS, Euro-Par, HiPC, ICPP,). He is currently the Belgian representative of the European Cost IC805 Action on Complex HPC. He is member of the Board and scientific advisor of CETIC (Applied Research Center in ICT, Charleroi, Belgium, co-founded by UMONS) and the current coordinator of Graduate School GrasCOMP in computer science, French Community of Belgium. He teaches courses in Parallel and Distributed Computing, Methodology of Programming, and Real Time Computing. He acts regularly as reviewer for Research Project Proposals or Follow up, either at the Belgian, French or European level.
Pierre Manneback has published more than 60 papers and communications at international conferences in the field of scientific computation and parallel computing. He is particularly interested in porting high demanding applications on new parallel architectures, either local (multicore, GPU) or distant (Clouds and Grids).
Optimizing Performance of Batch of Applications on Cloud Servers exploiting Multiple GPUs
Abstract:
Cloud computing is a computer paradigm aiming to provide remote and ubiquitous access to high processing power and huge storage at low price. Cloud servers are supposed to be heavily loaded in order to reduce usage cost. In this context, GPUs offer very interesting characteristics that can boost many applications. Integrated with Cloud computing, GPU usage can enable better energetic efficiency and lower environment impact. Our work is focused on the design of a management system to share multi-GPUs resources between many virtualized applications. This talk will focus on performance optimization when many applications are simultaneously loaded on servers exploiting GPUs.