340 likes | 429 Views
GRID: a research ‘product’ offering great opportunities for the industry. Federico Rossi Grid Business Development Director Casimir Workshop Milan - 24 February 2005. Outline. Grid: a short introduction stolen” ( with permission from Grid guru Prof. Domenico Laforenza CNR-ISTI)
E N D
GRID: a research ‘product’ offering great opportunitiesfor the industry Federico RossiGrid Business Development Director Casimir Workshop Milan - 24 February 2005
Outline • Grid: a short introduction • stolen” (with permission from Grid guru Prof. Domenico Laforenza CNR-ISTI) • Grid “for” the industry: the potential users • Grid “from” the industry: the solution providers • DATAMAT experience • How to build Grid future • Conclusions Casimir 24/02/2005
e-Science and Grid • The term e-Science refers to large scale research, carried out through distributedglobalcollaborations over the Internet. • The individual users need access to very large data collections, very large computing resources and high performance visualization • Source: http://www.rcuk.ac.uk/escience/ • The enabling technologies for e-science are: • Internet and the Web • allowing to discover and access information on a global scale; • DigitalLibraries • a new form of scholarly communication infrastructure that holds the promise of enabling more complete exploitation of knowledge and yields new capabilities and capacities so effectively and efficiently that they are intuitive and transparent in their operation (ubiquitous knowledge environment). • Grid • an infrastructure to share capabilities, to integrate services and resources, and to develop active collaborations across distributed, multi-organizational environments; . Casimir 24/02/2005
The Grid Utility Computing On Demand Computing Casimir 24/02/2005
Electrical Grid • Electric power applications have caused radical changes into the individual and collective life of human being Casimir 24/02/2005
The 1st Generation Grid Grid Evolution - Metacomputing • Different Supercomputing Resources • geographically distributed • used as a single powerful parallel machine (clear, High-Performance orientation) Casimir 24/02/2005
The 2nd Generation Grid Grid Evolution – Scalable Virtual Organizations • Grid computing has emerged as an important new field, distinguished from conventional distributed computing by its focus on large-scale resource sharing, innovative applications, and, in some cases, high-performance orientation. • The Anatomy of the Grid: Enabling Scalable Virtual Organizations Ian Foster, Carl Kesselman, and Steven Tuecke Fall 2001 Casimir 24/02/2005
Grid Evolution – Interoperability • But an open question remains: • is the far-reaching vision offered by Grid Computing obscured by the lack of interoperability standards among Grid technologies ? • (where interoperability describes whether or not two components of a system that were developed with different tools or different vendor products can work together) How to guarantee interoperability among Grids ? Casimir 24/02/2005
Grid Evolution – Interoperability • The marriage of the Web technology with the 2nd Generation Grid technology led to new and generic Grid Services • The Physiology of the Grid An Open Grid Services Architecture for Distributed Systems Integration: I. Foster, C. Kesselman, J. Nick, S. Tuecke, January, 2002 • http://www.globus.org/research/papers/ogsa.pdf GRID Computing + Web Services = Grid Services The 3rd Generation Grid Casimir 24/02/2005
Outline • Grid: a short introduction • Grid “for” the industry: the potential users • Early adopters • Target sectors • Grid “from” the industry: the solution providers • DATAMAT experience • How to build Grid future • Conclusions Casimir 24/02/2005
The dilemma of innovator and the ‘early adopters’ • The new technology changing the way user works can be considered: • Disruptive • An expensive toy • Innovation must be focused on the user operational context with his actual user needs. • To overcome the chasm between early adoptersand more pragmatists or conservativesusers, solutions and convenience must be shown beyond the satisfaction of basic needs. Casimir 24/02/2005
The “traditional” distributed supercomputing applications • High Performance Technical Computing for: • Automotive • Fluid-dynamics (e.g. Fluent) • Structural analysis (e.g. MSC-Nastran™) • Crash test • Chemical / Pharmaceutical • Bio-molecular simulation integratingexisting biological and chemical codes • Design of advanced materials (nano-, catalysts, … • Aerospace • Coupled simulations • EMC • CFD / Structural analysis Casimir 24/02/2005
The emerging ubiquitous networked society and e-Collaboration market • The information society is proceeding in the direction of both distribution and integration. • New value chains are being created, as appropriate for each situation, thus building “Virtual Organisations” characterised by being: • Objective driven • Time limited • Multi-enterprise • Dynamic changing the scope of traditional processes and workflows. • e-Collaboration offers an unprecedented way for organisations to facilitate coordination and communication, decision making and sharing of knowledge, skills and resources “Challenge: a common data standard for integration of operations" "In a recent Oil and Gas Industry Survey 84% of companies think e-collaboration is important." “Grid technology enables e-collaboration" Casimir 24/02/2005
Grid enabling e-collaboration: some examplesCollaborative / concurrent engineering • Aeronautics • Distributed design and testing(e.g. AIRBUS with production plants spread all over Europe) • Subcomponents / subcontractors management • Air Transport • Service level management / validation • Space • Distributed Concurrent Design Facility,extension of concurrent engineering approach to following development phases Casimir 24/02/2005
The new sectors being explored (1) • Finance • Risk management / portfolio evaluation • Health care • Surgical simulation • Large data bases Screening • Crisis management • Scientific-technical-operational collaboration • Intelligent DSS Casimir 24/02/2005
The new sectors being explored (2) • Services for SME • Use of shared resources intechnological districts (e.g. PEP) • Media, entertainment • Rendering, 3d animation • Telecommunications • Network routing optim. • 3G, B3G,… • and more… • e-government, e-learning • Meteo for Tourism , Sport, … Casimir 24/02/2005
Outline • Grid: a short introduction • Grid “for” the industry: the potential users • Grid “from” the industry: the solution providers • DATAMAT experience • How to build Grid future • Conclusions Casimir 24/02/2005
The solution providers • Hardware, Middleware, Software manufacturers • IBM, HP, NEC, Fujitsu, Intel, Sun, SGI … • UNIVA, Unicore, Platform, … • Nice, Gridxpert, Gridsystems, … • Oracle, Microsoft, SAP, … • IT Service providers (integration, applications,…) • SAS, DATAMAT, Atos Origin, … • Telecom Service providers • BT, Telecom Italia, Telefonica,… • The “Alliances” (more or less sponsored…) • Globus Alliance, EGA, … • The standardisation bodies with industrial participation • GGF, W3C, OASIS, IETF, DMTF, … Casimir 24/02/2005
DATAMAT profile • A top player in Italian ICT market with 30+ years experience: • ~1,500 young and highly qualified employees (end 2004) • 184 M€ Sales, 23.5 M€ EBITDA (FY2003) • Supply ICT solutions for end users in specific segments of our vertical reference markets: • Bank, Insurance and Finance; • Public Administration and Healthcare; • Telecommunications and Utilities; • Defence, Space & Environment; • Industry. through the delivery of: products, systems and integrated services based on specific functional and/or context requirements • Leveraging on: • Strong application expertise (knowledge of client primary processes) • Strong technical and technological competencies (“state-of-the-art”) • Management and organizational capabilities in managing complex and high economic value contractual responsibilities Casimir 24/02/2005
One of our key success factor: continuous R&D investments • More than 32,000 hours dedicated to R&D in 1Q04 (125,000 in 2003), approximately 5% of total hours Research Continuous participation with leadership roles in national and international research projects financed by public and private bodies (European Commission, NATO, Ministry of Scientific and Technologic Research, Ministry of Defence) Return in term of revenues and grants for €3.7m in FY03 Development Continuous launch of new products and upgrades Only a small part of development costs are capitalised, €2.1m in FY03 Capitalized costs (net of amortization) for €1.9m in balance sheet at 31.03.04 Strong combination of technological and process expertise Casimir 24/02/2005
SpaceGRID Why DATAMAT in Grid? • Our traditional focus is on innovation, with a long record of international R&D projects followed by successful commercial exploitation. • We perceived Grid as one of the most promising novelties for the ICT market, enabling (data- or computing- intensive) distributed applications, but also emerging context such as e-collaboration. • To acquire a specific knowledge of the technology, to understand its potential business return in terms of application and services deployment, also by applying technology transfer across our reference markets, we decided in 2000 to participate in the first large Grid Project in Europe … … and in the following years to other European outstanding initiatives: Casimir 24/02/2005
Which kind of collaboration with research institutions? • The collaboration with research institutions, in particular from the computer scientists within the High Energy Physics community (INFN is the foremost example!), resulted extremely profitable. • The cross-fertilization resulting from mixing different backgrounds allowed our teams to improve in terms of: • Flexibility • Teamwork • Openness to innovative ideas while being able to transfer our traditional development culture in terms of: • Structured approach • Consolidated methodologies for analysis and design • Tight project management (resource allocation, planning, …). Casimir 24/02/2005
“Grid-business” perspectives • DATAMAT is looking forward to three, potentially inclusive, business models: • Grid Consultancy • Provide expertise to end-users to grid-enable their applications • Grid Application Service Provider • Develop grid-enabled applications, maintain them, also taking in charge the needed logical and physical infrastructure and the relevant services • Grid Service Provider • Provide grid services to end-users (utility / on-demand paradigm) to satisfy their temporary requests for computing or storage power. • Furthermore, it is important to underline the potential role Grid can play as “enabling technology” for Virtual Organisations and Collaborative Environments. Casimir 24/02/2005
Outline • Grid: a short introduction • Grid “for” the industry: the potential users • Grid “from” the industry: the solution providers • DATAMAT experience • How to build Grid future • Conclusions Casimir 24/02/2005
VISIBILITY Linux Camera Phone Fusion Energy MMS Mobile Business Nano Technology Bluetooth SMS Internet eBook Gene Technology eBusiness Speach Recongnition DigiTV WLAN Web3D UMTS Quantum Computers Peak of inflated expectations Technology trigger Trough of disillusionment Slope of enlightenment Plateau of productivity MATURITY Which risks? • The easily perceivable risks, that could jeopardise the implementation (at least on a reasonable time-scale) of identified business perspectives, concern in particular: • The technology The architectures, the available implementations, the relevant standardsare not yet mature enough to have a kind of “grid-out-of-the-box” to be easily demonstrated to potential industrial customer • The market Hindering factors for an adequate maturation are: the poor diffusion of innovative concepts, the consequent little confidence in the technology as solution to real problems, a commercial context where often ‘launch’ takes place much earlier than actual availability, and the fragmentation of a pseudo-offer. But don’t forget legal and social aspects!! Casimir 24/02/2005
How to mitigate such risks? • DATAMAT position, from an industrial viewpoint, is that beyond the actions undertaken by companies (individually or in association), it is important to interact with the public component of research, in order to achieve stability and continuity, the two main factors really making possible the commercial exploitation of the Grid. • This can be achieved undertaking and securing initiatives at all levels (National, European, Global) in terms of: • Infrastructure deployment • Middleware standardisation • Application development Casimir 24/02/2005
A pan-European Grid infrastructure deployment • Relationship with public sector research is here fundamental to deploy and operate a wide and reliable Grid Infrastructure that can also be available, at least initially, to Industry. • This will enhance confidence in Grid from those end-user industries that have problems suitable to be solved with the help of Grid, but that still do not have enough evidence that such technology could give benefits on a large scale and in an industrial arena. • We consider a first important step in this direction the EC FP6 initiative EGEE, aiming at setting up a production-quality infrastructure for research: such effort should not stop with the end of the project, but should as well be supported as a long-lasting initiative. Casimir 24/02/2005
A convergence of middleware on Grid for business… • From an industrial viewpoint, it is mandatory that middleware technologically converges on solid business models and shared architectures to be exploited on Grid Infrastructures. • As Grid is about dynamic sharing and collaboration across multiple administrative domains, this means to enhance and standardise all technological aspects such as: • Security, e.g. Trustworthiness, Integrity, Confidentiality, Identity Identification, Privacy, Reliability, … • Service orientation, e.g. QoS, SLA, fault tolerance, … • Grid economics, e.g. Accounting, Payment, Compensation... plus • Certification (as a result of an industrial level engineering process) • In few words, to support a next generation of Grid technology to move the focus from e-Science to e-Business. Casimir 24/02/2005
… in a coordinated way • To ease sharing of developed solutions and technology take-up (in particular for SME’s), and start building value-added services, it makes sense that public and private research institutions commit themselves in setting up initiatives aimed at: • Promotion, integration, validation and packaging of platforms and solutions, • Repository of certified component / solutions • Education, Training, “after-sales” support • This would allow to complement on going initiatives, such as gLite development in EGEE, or next generation grid architecture definition in NextGRID, by long-lasting foundations, not just bounded to a single (or a series of) project(s). • It is reasonable to ensure it at the global level, taking into account, harmonising, and federating on-going efforts at continental and national level (e.g. US NMI, UK OMII, IT c-OMEGA,…). Casimir 24/02/2005
A new generation of Grid-enabled applications • The initiatives on application side shall: • Ease the diffusion of a Grid culture, at most through a generalised test-bed policy aimed at demonstrating in concrete terms the advantages of using it, in strong synergy with research • Grid enable classes of applications by developing suitable tool kits and/or porting existing problem solving environment on Grid contexts • Standardise (standardise, standardise, …) as much as possible the access to Grid services and resources, to strongly reduce the dependency on specific middleware implementation and/or specific infrastructure • Explore new working environments (e.g. ambientintelligence, nomadic / mobile environment) Casimir 24/02/2005
Grid in Lombardia and the Four Motors Initiative • Milan (Dept. of Physics and INFN) is one of the main centres of Grid development and deployment • Leading role in EGEE middleware and in exploitation by the applications (ATLAS experiment at LHC): 8 persons working with contracts funded by EGEE • The Representatives of the Regions which comprise the "Four Motors for Europe” (Baden-Württemberg, Catalunya, Lombardia, Rhône-Alpes ) have recently signed a • “COLLABORATION AGREEMENT TO SPONSOR COORDINATED DEVELOPMENT AND DEPLOYMENT OF AN INFORMATION GRID INFRASTRUCTURE FOR SCIENCE AND TECHNOLOGY AND ENHANCED SCIENTIFIC AND TECHNOLOGICAL APPLICATIONS” • The main goal is to spread and disseminate the culture of Information Grids, attracting in the initiative all the interested partners on the territory • The Lombardia Lead Partner for the GridPort is the University of Pavia Casimir 24/02/2005
Outline • Grid: a short introduction • Grid “for” the industry: the potential users • Grid “from” the industry: the solution providers • DATAMAT experience • How to build Grid future • Conclusions Casimir 24/02/2005
Conclusions • Considering growing interest of industrial user community (see enterprise grid days at GGF-12, EC FP6 ID’s, …), initial expectations on Grid are going to be fulfilled. • Initial enthusiasm is now being replaced by consciousness of the actual potential and risks, in terms of business and ROI, adequately steering strategic technical choices (e.g. web services-grid convergence). • From an industrial viewpoint, to make profit out of technology, it is fundamental to achieve the needed stability and continuity, focusing new initiatives on bridging research and industry. • DATAMAT, while confirming its commitment in on-going and future innovation activities, believes that Grid community should push stakeholders to: • Put in action a long-living global Grid infrastructures deployment plan (migrating towards new models, e.g. DANTE) • Ensure the support to the technological convergence of middleware on Grid for business • Put in place promotion, education, certification, repository, technology transfer activities to widen Grid culture diffusion and favour uptake in new application contexts. Casimir 24/02/2005
…and thanks to EC and ESA institutional websites and EGEE, EUROGRID, SIMDAT, CROSSGRID, SPACEGRID, GRIA, HEALTHGRID… project websites, useful sources of images and information MANY THANKS FOR YOUR KIND ATTENTION