DAILY NEWS AND INFORMATION FOR THE GLOBAL GRID COMMUNITY /    ( Table of Contents )

Special Features:

INTERVIEW WITH WOLFGANG GENTZSCH OF MCNC
By Alan Beck, Editor-In-Chief

ALAN BECK: Why did you decide to leave Sun for MCNC?

WOLFGANG GENTZSCH: I achieved my goals at Sun. When Sun acquired my company, Gridware, in 2000, my three main objectives were to help integrate our team and software into Sun, to make the Sun Grid Engine distributed resources management technology ubiquitous, and to evangelize Sun's Grid computing vision within the organization and to the rest of the world.

After more than three years, I consider all these goals successfully achieved. Grid Engine is the most widely used Grid software for managing departmental and enterprise Grids and Sun is widely recognized as one of the global leaders in Grid computing. The time has come for me to move to the next logical step in my career. At Sun, we developed technology for department and enterprise Grids. Now, I will work with the team at MCNC to apply and deploy Grid technology to build turnkey Grid solutions for customers.

MCNC has provided a once-in-a-lifetime opportunity. MCNC is in the heart of North Carolina's Research Triangle Park and has an established reputation for its innovative contributions to networking, supercomputing and Grid technologies. With its wealth of experience, expertise and existing assets, MCNC is a natural place for developing, implementing and applying Grid technology.

In addition, North Carolina's state government, the public and private universities in the state, and much of the local industry are already aware of the benefits of Grid computing and are all very open to learning and deploying Grid technologies.

AB: How would you characterize your experience at Sun?

WG: As a technologist, you are coming to a great place when you join a company like Sun, with some of the brightest people in the IT industry. Grid computing was always supported by Sun's executive team. Indeed, internally, we reshaped Sun CEO Scott McNealy's famous quote, "The network is the computer," into, "The Grid is the computer." And with recently adding low-end Linux and Intel x86, we were able to offer heterogeneous Grid-enabled hardware and software solutions for the datacenter.

AB: Could you help us understand MCNC's business?

WG: MCNC is a private, independent, non-profit corporation established in 1980 to advance technology-led economic development and job creation throughout North Carolina. MCNC Grid Computing & Networking Services is building one of the country's first statewide Grid computing networks for research and education, and it delivers advanced communications resources across North Carolina to public and private institutions, including universities, community colleges, K-12 schools, libraries, state government, private research institutions and commercial businesses. MCNC Research & Development Institute develops new technologies through its own initiatives and as a research partner for businesses and the U.S. government, conducting advanced and applied research across a broad technology spectrum, including microsystems, flexible electronics, sensor development, signal electronics, wireless systems, microfabrication, high-speed secure networks and Grid computing. MCNC Ventures provides early-stage funding and assistance to entrepreneurial start-up companies.

AB: You mentioned MCNC's existing resources in general. What are some of MCNC's Grid computing assets?

WG: That's already a long list today. MCNC, in partnership with North Carolina's universities, has already launched a multi-year, multi-million dollar North Carolina Grid Initiative to add a Grid computing infrastructure and resources to its existing North Carolina Research and Education Network (NCREN). This innovative statewide network provides advanced communications services, video services for distance learning, high-speed Internet access, and access to national research networks for public and private universities throughout North Carolina.

Having a statewide network already in place is a key competitive advantage. NCREN is the backbone infrastructure for the North Carolina Grid Initiative and will serve as a reference implementation for commercial use of Grid computing for all of its existing and future partners and customers.

MCNC and North Carolina universities demonstrated national leadership in deploying computing Grids when it launched the North Carolina BioGrid in 2001 -- one of the nation's first Grid test beds for life sciences research. The Grid test bed has evolved to address applications in multiple disciplines with the launch last year of the MCNC Enterprise Grid.

As early adopters, MCNC is already addressing the challenges in deploying, operating, and scaling a production Grid infrastructure including information security and retrieval, joint collaboration, applications, middleware and network infrastructure provisioning. MCNC is also an active participant in standards development through the Global Grid Forum.

All these elements make MCNC an ideal place for me to continue working with Grid technology in an environment that spans research, development and deployment of Grid technology using the statewide research and education network.

AB: What do you hope to accomplish at MCNC?

WG: Today, companies are deploying Grid computing within a single department or enterprise, pooling computing resources into one virtual system. To realize the full potential of Grid computing in the future, resources will be linked across organizational boundaries for collaboration, efficiency and cost savings. Just as academia and the research community led and advanced the development of the Internet, MCNC and North Carolina universities have the opportunity to help pioneer Grid computing.

Our goal is to play a very active role in Grid computing for research, education and industry in North Carolina and throughout the United States. My personal contribution will be to work with the MCNC team to find out what our customers really need, and to design and implement Grid solutions to meet their needs.

North Carolina's Grid Initiative will be a catalyst for economic expansion and business growth. It is exciting to lead an initiative that will not only impact the future of computing but will play such an important role in North Carolina's economic advancement.

AB: What are the greatest challenges to achieving your goals?

WG: Our greatest challenge, as I see it, is to build the necessary relationships, trust and credibility in the application of Grid computing solutions so that our customers join us in the effort to learn about Grid computing. Together, we want to help improve our customers' IT infrastructure in strategic and manageable steps, and in a non-disruptive manner.

There are certainly other organizational and technology challenges as well, including security, identity, network bandwidth and resource sharing, but we are optimistic and confident that we can solve these challenges, one step at a time, by working with our customers and the Grid community.

AB: How do you see the state of Grid technology today?

WG: The current state of Grid technology, in my view, has two main components. One is the great and very ambitious effort to develop standards and technologies to enable virtual organizations, over the Internet, to access distributed resources for computing, communication and collaboration, in a secure and reliable manner. The Open Grid Service Architecture of the Global Grid Forum was a big step forward, but we are certainly still very much in a prototype and experimentation phase.

On the other hand, there is a more pragmatic approach of a few software companies (including Avaki, Datasynapse, GridXpert, Platform Computing, United Devices and Oracle) and some IT infrastructure players (including Hewlett-Packard, IBM and Sun) to develop products for departmental and enterprise Grids, for the datacenter, for the adaptive enterprise, for computing on demand, and so on. Here, partial solutions are already available and working. However, here too, the long-term vision of computing as a utility, even restricted to the enterprise, is still many years away.

AB: What obstacles stand in the way of widespread deployment of commercial Grid applications?

WG: It is my experience that IT managers today are largely not yet convinced that Grid computing delivers on its broad range of promises. They have a high level of accountability and responsibility to maintain around the clock service for their company. IT has become more complex and volatile over time, and there is great resistance today to embrace major changes unless you can demonstrate an order of magnitude improvement in productivity, utilization, reduction of cost and complexity, time to market, efficiency, etc.

Mostly in the area of so-called Global Grids, those which are reaching beyond firewalls, the main obstacles I see today are lack of widely accepted standards, immature and incomplete technologies, and only a few convincing use cases which demonstrate the great benefits a global Grid can provide for global businesses.

AB: What are the best approaches for overcoming these barriers?

WG: For IT within the enterprise, one of the best approaches is to introduce Grid technologies in small steps, in an evolutionary way. Start with your people; ask who should become our Grid expert? Then look at your existing environment; determine what systems can easily be Grid-enabled such that they become a part of the departmental or enterprise Grid, and what applications are ready for the Grid. If you are in the course of replacing or adding systems anyway, evaluate whether there are Grid-ready systems from vendors that fit the organization's needs, such as Grid racks or Grid blades with built-in Grid software for Grid system and workload management. The process should include evaluation about whether the technology is scalable -- whether it can manage, over time, thousands of nodes and hundreds of thousands of jobs or tasks. Then, after some years in production, there might be the growing need to connect your enterprise Grid to other Grids, to enhance your global flexibility, to interconnect your worldwide subsidiaries, or to better interoperate with your partners, suppliers and customers.

AB: How will MCNC help overcome those obstacles?

WG: We will create an environment that will allow us to work with customers to collaborate and learn about these challenges. We will continue our research activities and involvement in the standards bodies to develop solutions to challenges associated with moving Grid technologies from the research lab to production environments. In a collaborative environment, we will work with partners and customers for application benchmarking, interoperability verification, systems integration and operational training.