Some of the Points Gerry and Craig Discussed

— Despite the recent economic downturn, we are living in an era of digital prosperity. As the songwriter Paul Simon says, we are living in the age of miracles.

— But there are signs of trouble ahead caused by the input and output of computers, electricity and heat.

It's analogous to automobiles. In that case we don't have enough oil and we produce too much carbon monoxide. With computers we're struggling to provide enough electricity and to rid the machines of heat.

  What we need of new computing architectures. Just as a Prius or a Tesla is a new type of automobile, at Purdue we are using new — somewhat experimental — computers that are better at managing resources. This reduces costs and allows us to continue to do more work with flat or declining budgets.

  —If we don't respond to the challenges of not enough electricity and too much heat, we'll reach a limit on what we can achieve, which would be an anchor on the economy.

This would range from the socially important — constraining medical research — to the socially social, such as a limit on the number of photos you can put on your Facebook page.

—One approach, which is particularly timely in the current recession, is to get more computing from the machines you already own.

At Purdue we have about 35,000 computers. These computers are in actual use only about half the time at best. This is the story in every university and business in the United States.

What we've been doing differently for the past couple of years is putting the machine to work when the users isn't asking it to do a task. In 2008 we captured 16 million computing hours this way in 2008.

To put this in perspective, this would be the equivalent of a cluster supercomputer that would cost more than $3 million, require 2,000 square feet of floor space, and rank among the top 100 supercomputers worldwide.

  This is all from machines that we already owned that weren't doing anything.

  Any university or business should learn to do this in order to maximize their resources. People are looking past this technology because it's not sexy. But distributed computing is the future, whether it's sexy or not.

    —One coming technique is to use highly parallelized, high-throughput computing by utilizing grid computing or a specialized supercomputer such as the SiCortex Supercomputer, which is built near Boston.

Here's one catch—almost none of our computer code is written is such a way that it can be used on these machines or on distributed computer grids.

  Using this type of computing is as obvious as designing fuel-efficient cars. But we need programmers who can write parallel computing code (which is much more difficult to write) and engineers who can design these new types of machines. At Purdue we've jumped on this early and we are both designing computers and teaching students for these new computers.

    —Another approach is to use different chips that are optimized to do different tasks (an approach seen in the new Apple laptop computers).

  These machines are called hybrid computers, and they assign different computing jobs to the type of chip that can most efficiently. This is still somewhat in the research stage. Like the other types of computing I've mentioned, this requires a new way of programming the computer, which requires a work force with a different set of skills.  

  —A different story is in the power of social media.

Facebook and Twitter are hugely successful, but we've been looking at these for a couple of years and thinking, what if we took scientific communities, empowered them with computing resources, and then let them solve problems together?  

We call these "hubs," and in a couple of cases the results have been dramatic. For example, in this case, using nanotechnology to design future computer chips. This hub has more than 90,000 users, which has to be nearly the entire nano-electronics community.

  Instead of asking "Which Saved by the Bell" character are you, the scientists can ask each other questions. Here's one from this week that began: " I'm simulating a simple metal-silicon-silicon oxide-silicon-metal system, and everything seems to be working except when I try to add surface states at the silicon-silicon oxide border none of the results change…."  

I think we've all been there.

  But the scientists can use the community, crowd sourcing is the current buzz word, to advance their research.

  This is going on in a dozen or so scientific disciplines, and we expect to see rapid results from these connections.

About Gerry

Gerry McCartney was appointed Purdue's vice president for information technology and chief information officer in July 2007. He had served as the University's top information technology administrator in an interim capacity since July 2006.

As CIO, McCartney is responsible for overseeing Information Technology at Purdue, informally known as ITaP, and reports jointly to Provost Randy Woodson, and Senior Vice President and Treasurer Al Diaz. ITaP is responsible for the planning and coordination of central computing and telecommunications systems on the West Lafayette campus, along with media production and distance-learning services. The organization consists of six business units and has more than 450 staff members and an annual budget of more than $63 million. McCartney also holds an appointment as associate professor in Purdue's School of Technology.

Prior to his interim appointment, McCartney served two years as assistant dean for technology at Purdue's Krannert School of Management where he taught in the Krannert MBA, executive programs and engineering management program; he continues to act as principal investigator for the SIFT Project.

From 1993 until 2004, McCartney was associate dean and chief information officer at the University of Pennsylvania's Wharton School. He was director of the Krannert Computing Center from 1990 to 1993 and was manager of user services at the Purdue Computing Center from 1988 to 1990. He has also held managerial positions in the computing centers at both the University of Notre Dame and Maynooth College in Ireland. He holds a patent, was awarded the CIO Enterprise Value Award in 2003, and is a member of the CIO Executive Council. He speaks and comments frequently on the entrepreneurial management of technology and is a contributor to Cranky Geeks, an Internet show focused on technology.

McCartney earned his doctorate in sociology and anthropology from Purdue in 1995 and diplomas in advanced computer programming and systems analysis from the Graduate School of Engineering at Trinity College in Dublin, Ireland, in 1982 and 1984. He took first class honors in both his bachelor's and master's degrees in 1981 and 1984, respectively, from Maynooth College.

Gerry is married to Dr. Kitti Carriker and they have two sons, one attending Purdue and the other attending West Lafayette High School.

About Purdue

Purdue University is a coeducational, state-assisted system in Indiana. Founded in 1869 and named after benefactor John Purdue, the University is one of the nation's leading research institutions with a reputation for excellent and affordable education.

Building upon historical strengths in engineering and agriculture, the West Lafayette campus currently offers 5,800 courses in more than 400 undergraduate majors and specializations in the schools/colleges of Agriculture, Consumer and Family Sciences, Education, Engineering, Health Sciences, Liberal Arts, Management, Nursing, Pharmacy and Pharmacal Sciences, Science, Technology, and Veterinary Medicine. Programs of graduate study and research leading to advanced degrees fall under the jurisdiction of the Graduate School.