Posted by Dan Reed
Corporate Vice President, eXtreme Computing Group and Technology Strategy & Policy
Throughout the history of science, back to the days of the Renaissance, data has been scarce and precious. But today, riding the same technological economics that have given us inexpensive computing and ubiquitous sensors, scientists have the ability to capture data at a previously unimaginable scale. In all domains, scientists and researchers are drowning in data. They've gone from scarcity to an incredible richness, necessitating a sea change in how they extract insight from all this data.
In a parallel shift, our scientific questions and problems increasingly lie at the intersections of traditional disciplines—for example, the recent U.S. oil spill in the Gulf. Understanding the complexities of what it means for oil distribution in water is a problem related to computational fluid dynamics, but understanding the impact of that oil on the marine ecosystem is a biological problem. To fully understand the issue, researchers from multiple disciplines—from different cultures, using different research tools—have to unite to build models and analyze data from diverse sources.
With this has come an insatiable demand for easy-to-use tools and computing support, unfortunately requiring many researchers to assume additional systems administrator roles. These researchers often spend inordinate amounts of time maintaining the computing systems they require to do their research rather than devoting their time and talents to the research itself. The cost to maintain and refresh this computing infrastructure is becoming a larger and larger burden, and the economics are unsustainable. As a result, much of our research funding climate has focused (because of the power of computing for scientific discovery) on refreshes and repeated deployments of infrastructures on research campuses and laboratories. Yet at even the best funded research organizations, the majority of researchers do not have access to the computing resources they need.
Fortunately, the emergence of cloud computing coupled with powerful software on clients, such as a local desktop computer, offers a solution to this conundrum.
Cloud computing can provide software applications and computing power to users as a service over the Internet via familiar tools. The offsite cloud is constantly managed and upgraded, providing the ability to deliver computational resources on demand, a “pay as you go” strategy with access to large scale computational capacity. The cost to use 10,000 processors for an hour is the same as using ten processors for 1,000 hours, but will deliver radically faster analysis to the researcher. Organizations can buy just-in-time services to process and exploit data, rather than on infrastructure.
Further, the cloud offers unique opportunities to support a global, multi-party and neutral type of collaboration—allowing a diverse set of experts scattered across multiple continents to bring their expertise to bear. For example, in the future of drug discovery, our desire is to develop drugs that are targeted to individuals, not broad populations. To do so, the expertise of large pharmaceutical companies, the basic research of academia and the growing body of medical knowledge and skills around the world all need to come together in a seamless way. There are a whole host of issues that come with this—extracting the relevant data, correlating concepts, bridging cultural and technological divides, alleviating competitive concerns and so on. The cloud allows all these parties to access the data in neutral ways in their own context via their own familiar tools and collaborate using many different models and designs, simulations and experiments.
By extending the capabilities of powerful, easy-to-use PC, Web, and mobile applications through on-demand cloud services, the capabilities of the entire research community will be broadened significantly, accelerating the pace of engineering and scientific discovery. The net effect will be the democratization of research capabilities that are now available only to the most elite scientists.
To accelerate this paradigm shift, we’ve launched a worldwide initiative to work with governments and academic institutions to explore the power of rich client tools coupled with powerful cloud computation and data storage. The objective of the program is to provide massively scalable tools and services directly to researchers that they can access from their desktops using familiar software tools. Microsoft will work with research organizations to provide free access to advanced client-plus-cloud computing, the tools to access the cloud, and important cloud services and technical support. The research organizations will identify the programs that will receive the free cloud services and support.
To date, Microsoft has signed agreements with the U.S. National Science Foundation, Japan’s NII Info-Plosion project, and today, in Europe with the European Commission’s VENUS-C project, France’s INRIA, and the University of Nottingham.
We believe our efforts can help to drive a global shift—a shift to invest more in research, in the acquisition of cloud services rather than in the distributed maintenance of infrastructure, allowing researchers to focus on unsolved questions and discovery, not on computer systems administration. If you’re interested in reading more about our vision of how client plus cloud computing can democratize research, see our recently published White Paper here.