Find the most recent information on EU Funding activities in the field of Information and Communication Technologies (ICT) by visiting our ICT in FP7 website , which covers ICT in the 7 th Framework Programme (FP7) 2007 - 2013.
IST 2003 Conference (Grid-related events in the Conference Programme)
: 2-4 October 2003
Location : Milan, Italy
Report of the Workshop " Grid Economy and Business Models "
(part of the Spotlighted Topic "Harnessing computing and knowledge resources")
Grid technologies offer the promise of new productivity gains in the future knowledge-based economy. Grids will foster new business value chains and boost innovation capabilities by supporting the creation of collaborative working environments across multidisciplinary scientific, engineering and industrial communities. This session addressed cross-domain issues related to the creation and management of Grid-enabled virtual enterprises, business models for Grid services, Grid economies and market mechanisms for eBusiness co-operation.
The session chair welcomed the audience to this discussion on the Grid economy and business models. Grid computing offers advanced capability for enhanced access to vast amounts of computing power and data and the session reviewed progress from the present state-of-the-art in academic and industry to the formation of a Grid economy, that is to say, a set of marketplaces comprising buyers, sellers and services. The first two speakers discussed technical matters and the second two discussed developments leading to vital business models. The presentations covered the identification of potential users outside academics, major industries and governments, and examined if SMEs or regional hubs could use Grids. The session was directed at raising a whole set of issues in the hope of generating food for thought and new ideas and opportunities.
Main Issues Raised
Dr. Rajkumar Buyya spoke first about how the Grid would be delivered and what challenges must be overcome. Some design parameters that must be considered for the Grid economy are its heterogeneous and distributed nature and the wide range of interest groups. Mechanisms are required for value expression, translation and enforcement such as how to price services. Dr. Buyya then proposed a reference architecture incorporating open-source software technologies (from the Gridbus project), a Grid architecture stack for running a Grid economic model, Grid workflow, a Grid resource broker and a Grid portal.
Drawing from a case study of drug discovery using Grid resources to screen millions of compounds to identify drug candidates, Dr. Buyya mentioned the experiment set-up of 165 jobs each consuming five minutes of Central Processing Unit time. This system selected resources from a global pool at a negotiated cost per Central Processing Unit second, thus keeping the budget in control. In summary there are open research opportunities and challenging issues like dynamic pricing and protocols for service negotiation to be determined before a Grid marketplace could be built.
Dr. Steven Newhouse presented work from the London e-Science centre that aims to build a production Grid for United Kingdom science and business with funding from the United Kingdom Government. Dr. Newhouse observed that a lot of Grid activity was uncoordinated at a technical level thus making it hard for business to get involved. A big effort on standardisation is required and this is the barrier to building Grid economies and systems and making the Grid work. The Global Grid forum has been established to progress these issues in a format similar to the Internet Engineering Task Force (IETF). The forum is developing Grid implementation standards, some of which leverage the Internet, and these will offer the Grid community the common nomenclature and semantics required to form an interoperable open Grid services architecture with a defined core infrastructure that services can plug into. This will enable truly virtualised resources to interoperate using service discovery, job execution, pricing and brokering.
The United Kingdom Grid Marketplace project is focussed on business applications of Grids in partnership with some major corporations and Information Technology suppliers. It has been running for six months and is presently working to understand business requirements such as trust, reputation and reliability for service discrimination. The project has already demonstrated a basic implantation and contributed to standards work. One major challenge is to replicate price discrimination after the model of the real economy on such issues as how good a service is and what interests users have to use it. This requires some form of brokerage as part of the service presentation that can represent many different access models including fixed pricing, academic discount and other open market mechanisms. The service's internal accounting needs to use existing mechanisms such as credit cards, and in future eCash. When standards are established a large number of small-scale services will emerge and a mechanism will be needed to locate services from around the globe. The Grid will provide many opportunities that companies can exploit as content providers, service providers and intermediaries.
Dr. Andrea Nicolai gave a presentation of digital ecosystems and new software service paradigms with a vision from the business side drawing from the waves of Internet growth with many types of networks, services, protocols and standards. Like the Internet, the Grid system will be a wealth generator and to succeed it must offer an open infrastructure that can cope with heterogeneous, unreliable and insecure networks, variable latency and bandwidth, evolving topology and multiple administration domains. Successful business applications using Grid computing will expect an environment that is open, reconfigurable, lightweight, semi-transparent, user friendly and reliable. Also, there needs to be a service-to-service model whereby clients can reach each other without intermediate servers. These Grid services must be direct, immediate and offer high efficiency, be more performant, and be driven by economic considerations.
Successful Grid computing has the potential to simplify Information Technology services in large corporations by turning Information Technology into a variable cost where customers pay only for what they use. This can significantly reduce operating costs. For this selective outsourcing to work some critical conditions must be fulfilled such as open access, open- source and open standards that are royalty free, working seamlessly within a regulatory framework with mutual trust. Dr. Nicolai believed the Grid's weakest links are presently language and definitions. To move forward from the business perspective there has to be a symbiotic architecture and language so that self organising enterprises can make a digital representation of their needs to translate them into self co-ordinating service components embedding business logics in a digital ecosystem . Research activities in this Digital Business Ecosystem should facilitate wider adoption of Information and Communication Technologies and support SME activities to cross the digital divide by placing dynamic services on top of open-source infrastructure or commons. However, this work must always keep in mind that the network will tend to be un-controllable because architecture is often driven from the edge once it places more power in user hands, leading to the creation of so-called digital Darwinism - a service that is able to evolve as users need it to.
Finally Dr. Marko Seppd gave a presentation of future Grid-oriented activities in the Tampere region of Finland and the e-Tampere programme. Tampere is becoming a high-technology region where Grids will be used by academics, businesses and government. Dr. Seppd discussed how Grid technology would contribute to major societal change in which capitalists could be replaced by knowledgists as the key enablers of the economy. There was evidence of this big change taking place that was termed knowledgisation , that is to say, the application of human capital. One way to achieve this is to foster the Grid economy and there was fresh evidence from the IST 2003 event that value networks may replace the traditional firm. Dr. Seppd went on to describe the major industries in Tampere area and its economic development as traditional industry has evolved to high technology, including the impact of Nokia, and further efforts to boost Information and Communication Technology development. Tampere is operating a local pilot of e-Europe that makes up e-Tampere and there is a new technology company incubator called the e-Accelerator. On-line services for citizens are available through the InfoCity service and there are emerging signs of success and wealth creation. More experience of Grids was required in businesses, universities and government but there are already early successes paving the way to the digital economy.
Conclusions and Future Directions
Discussion between the audience and the speakers opened with a question regarding where the line was between open-source and business model applications that can drive industry and performance and whether any consensus was emerging. The speakers confirmed their opinions that the Grid's enabling infrastructure must be open and, like the Internet, based on open/free software such as Linux, the Apache web server, various web browsers and Java clients. Probably the lower levels will be open-source but at the high level there will be more closed source code, which remains standards compliant.
The session chair asked whom the first Grid traders might be with early adopters and customers who will buy services. Dr. Newhouse's opinion was this could start quite soon in the United Kingdom academic community and it would be interesting to see how introducing an economic layer would change the way it was used. He expected to see Grid powerhouses that sold computing and storage to grant holders that were funded by the research councils. It would take investment to encourage the business community to share standards and semantics and it was imperative that standards were created for the future thus avoiding barriers. There is a lot of work on the right track and standardisation is biggest issue.
Another question resulted in discussion about the importance of the ability to negotiate and make trade agreements and how to advance existing standards through making a smooth evolution in addition to vision. Engagement in the open standards process at business forums was seen as important and a first step is for businesses to pool knowledge about each other's activities and requirements. There is a series of meetings in the United Kingdom defining a new protocol and framework for providers and consumers to exchange information on their needs, negotiate and reach agreement. All business partners and standards setters are engaged and it is hoped that these standards will converge towards something everyone can live with.
The session chairman concluded there are viable proposals for architectures that would generate Grid economies of which some were already implemented and used in pilot projects to see how the Grid may operate. There were many potential players in the Grid field with sufficient long-term vision from business about how to energise and participate in these new opportunities. Lots of open questions remain regarding how business will take up Grid technology created by computer scientists, but the session had delivered an encouraging message.
For further information (including the possibility to download presentations given at the workshop) visit the workshop's page on the conference website.
- Session Chair
- Professor Christos Nikolaou , University of Crete, Greece
- Dr. Rajkumar Buyya , GRIDS Lab, University of Melbourne
- Marko Seppa , Director e-Business Research Centre, Tampere University of Technology
- Andrea Nicolai , Chief Executive Officer T6, Italy
- Steven J. Newhouse , London e-Science Centre, Imperial College
- Commission Contact
- Vincent Obozinski