Technology and knowledge transfer are being emphasized more and more in publicly funded research. Distributed computing actually has a long history of successfully transferring technology that benefits the global economy, but this has sometimes been overlooked.
In times of global economic recession, governments want to know that, when reaching into the public purse, they're getting the maximum possible benefits from the money they give to scientific research. They're looking for value and increased economic growth, driven by innovation. E-infrastructures, and the research they underpin, represent some of the most ambitious projects in modern science, both in terms of scope and the resources required to carry them out successfully. There is little doubt that the amount of public money spent on e-infrastructure projects is being closely scrutinized by economists and politicians alike. As such, it is imperative for research institutes to make links with industry and show they can contribute to economic success, both on national and international scales.
Scientific computation, however, has always maintained strong links with industry, especially compared with many other research disciplines that are equally reliant upon public funding. One of the primary reasons for this is the field's dependence on the development and release cycles of various computer manufacturers, who in turn listen to the research community and compete for their business. By demanding computers with ever-greater power to solve bigger problems more efficiently, the research community also plays a role in creating a trickle down effect: the first forays into parallelization of computational problems in the 1980s have impacted on the development of multicore mobile devices today, for instance.
Scientific computation is also a discipline that is very effective at producing researchers who are comfortable working in industry. The success of large-scale e-infrastructure projects depends on collaborators who are "adept at solving complex issues… but also good communicators", says Silvia Olabarriaga of the Academic Health Centre in Amsterdam, The Netherlands. Olabarriaga has coined the term 'peopleware' to describe individuals who develop highly transferable skills through working in international scientific computation projects.
The flow of knowledge out of publicly funded research need not just depend on highly trained personnel leaving the hallowed university halls or shiny research centers for jobs in the private sector, however. Ian Osborne of the Knowledge Transfer Network in the UK feels that making use of innovations developed in academia through the right channels can be crucial for small-to-medium-sized enterprises: "Small companies can use licensed technologies developed in academia to help fulfill their product pipeline without so much investment in R&D," he said.
Scientists are used to dealing with big data, and yet the magnitude of the data being generated is growing at a seemingly exponential rate, leading many experts to forecast a looming 'data deluge'. Tony Hey, corporate vice president for technical computing at Microsoft and former head of the UK e-science program, thinks that not only scientific researchers, but also commercial enterprises will actually benefit from learning how to deal with big data. "The magnitude of the data and the complexity of the analytical computations will drive both scientists and business analysts to explore the use of cloud computing, high performance computing, multicore processing and parallel programming to manage the computing workload," he says.
Transfer of technology and knowledge is, of course, a two-way street. This is true for computational science, as well as other research disciplines. Consider, for example, general purpose computing on graphics processing units, or indeed the invention of the World Wide Web itself. Perhaps the greatest impact of technology and knowledge transfer is ultimately the way the drive for better, more efficient technology improves all our lives. The smartphones of today, for instance, challenge the supercomputers of just two to three decades ago for raw speed thanks to system-on-chip integration, but more importantly they also connect people around the world to the web for the first time, vastly increasing human potential.
This article was adapted from the latest e-Science Briefing on technology and knowledge transfer.