Editor's note: Last week, President Bill Clinton named Ken Kennedy, the
Noah Harding Professor of Mathematics in the Department of Computer Science,
as co-chair of the new federal advisory committee on High Performance
Computing and Communications, Information Technology and the Next Generation
Internet. Kennedy, who is also director of the Center for Research on
Parallel Computation, an NSF science and technology center based at Rice, is
one of 20 representatives from academia, industry and government invited to
serve on the committee. His co-chair, a representative from industry, will
be named at a later date.
"President Clinton could not have made a wiser choice than Ken Kennedy,"
said Rice President Malcolm Gillis. "Ken will bring to the advisory
committee a wealth of experience in high-performance computing and a bold
vision for the future of computers. Ken's direction of the CRPC at Rice has
been an indispensable element of our initiatives in computational
engineering. We can expect that, under his leadership, the committee will
achieve its ambitious goals."
Kennedy talked with News Office Director Michael Cinelli about the
issues facing committee members as they prepare to meet for the first time.
Michael Cinelli: The Next Generation Internet (NGI) initiative announced
last fall by the Clinton Administration will be one of the first topics for
the advisory committee to tackle. With all the concern about how the current
Internet is clogged with "junk," is it time for the next generation or
should the current generation get cleaned up first?
Ken Kennedy: Next Generation Internet will not solve the problem of the
first generation being clogged with "junk." If anything, the problems we're
experiencing with the first generation will get worse. The solution to those
problems will be in the development of additional software and policies to
filter out much of the "junk" that is clogging up the system. Technology
will continue to improve and that should resolve the situation.
Those problems, however, should not hold us back on getting a jumpstart
on developing applications for use with the NGI. The next generation
will take us a step beyond where we are now. It won't just speed up access
to the World Wide Web. It will enable many applications that are difficult
if not impossible to run on the current Internet. There will be a lot more
full-motion video, greater use of supercomputer applications, plus we will
see high-quality interconnections to high-quality interfaces. There will
just be an enormous increase in real distributed applications that will run
on computers around the country.
The first role of NGI will be to get the research and development
community started working on the types of applications that will make the
Next Generation Internet a more effective tool for the average person.
MC: How will the committee address the debate about how computers and
high technology will be available only to those who have computers, thus
creating a new class of poor and unconnected citizens around the world?
KK: I don't think the cost of computers will be impacted that much by
the NGI. Most people will still be connected by modest speeds using their
personal computers. Connections for the average person will continue to
improve at the same or lower prices. And we will find more economical ways
to communicate on the Internet, for example, through Web TV.
Clearly the goal of the Clinton Administration is to make sure that at
the high end of the spectrum, the people who need to have access to the Next
Generation Internet are granted it in an equitable way that does not
eliminate whole communities that could benefit from it. For instance, K-12
education could benefit from NGI.
So, I am sure the committee will talk about how to make certain
affordable connections are available for everybody. The cost of an NGI
connection for a university will be substantial. But those are issues we
will talk about: how to leverage federal investments to make NGI accessible
to everyone. We will, however, stay away from any regulatory issues.
We will be more interested in how to direct federal investment in order
to get the maximum impact from research and development of NGI to the
taxpayers. We will discuss who should pay for what, how much special
investment of moneys should go toward K-12, how local school districts and
the university research community should be connected, how much federal
investment should go toward those issues and how much should come in the
form of a discount from telecommunication companies.
The telecommunication industry can't pay the whole bill without going
broke, so there has to be some equitable distribution of cost. Everyone has
to pay some part of the bill, including the federal government.
MC: Immediate access to information has been the hallmark of first
generation Internet. What frontier will the Next Generation Internet explore
and possibly conquer?
KK: High bandwidth applications are the key to that. There will be some
improvement in the immediacy of access, but having high bandwidth in the NGI
will provide better service for some kinds of requests that require a lot of
downloading of information. The time to access the server, however, will not
improve much. Downloading big files will go much faster.
And since you will be able to download big files, things such as fullmotion
video will become possible. Currently there are web sites with
streaming video, but that is extremely primitive. Many new applications have
yet to be designed, but soon we will be able to link all the databases in
the world so we can get regular, tailored updates on information in ways
that are not possible today.
We will be distributing computer software systems to run on global
computer networks so that visualizations can be provided at the high
bandwidth of the NGI. That will mean, for example, that product designers
will be able to sit at workstations in different locations and share
computer design simulations. They will have access to up-to-date parts lists
for trying alternative designs, providing the NGI can be made sufficiently
secure.
MC: How will the committee address the security issue for NGI?
KK: Speaking as co-chair of the committee, one of our goals will have to
be to address those problems adequately, giving connections to the research
and development community high priority. We will advise the federal
government that it should invest in research to develop reliable and secure
computing networks.
One of the initiatives the federal government is starting now, with or
without the Next Generation Internet, is in high-confidence computer systems
which employ cryptography in new ways to achieve higher levels of security.
There will be a lot of activity in this area for a long time to come.
The other side of the coin is making certain there is high reliability
in the NGI. We have achieved that on the existing Internet. One line goes
down but the Internet still functions.
The key issue facing the committee will be how to get this thing (NGI)
started for research and development at universities and companies and what
is needed to make it a useful resource for the country.