Panel Will Urge Major Spending for Computer Research
From: New York Times, July 27, 1998
By John Markoff
SAN FRANCISCO -- A presidential advisory commission made up of some of the nation's leading computer scientists plans to deliver a report Monday to the White House, calling for a significant increase in federal spending on advanced computer research in order to create "expeditions to the 21st century."
The report, a draft of which was given to The New York Times by someone close to the committee, calls current federal support for research in information technology "dangerously inadequate," and too narrowly focused on "near-term problems."
The panel, known as the Presidential Advisory Committee on High Performance Computing and Communications, Information Technology and the Next Generation Internet, has as its co-chairmen William Joy, a founder of Sun Microsystems, and Ken Kennedy, a computer scientist at Rice University.
The draft report calls on the federal government to finance research centers that will pursue broad-ranging computer research in the spirit of the early explorers of North America.
"Much like Lewis and Clark opened the West," it says, "virtual centers focused on future technologies and applications can, by making bold assumptions about the future, give us key research insights into the manifold possibilities of the 21st century technologies."
Such new research centers might build upon, but move beyond work done at today's federally financed supercomputer research centers and their attempts to advance scientific computing.
Computer research with such government backing has resulted in some of the most significant new industries created since World War II, including personal computing and the Internet. Almost all of those advances came from basic exploratory research, rather than research and development efforts aimed at solving specific problems.
"Although total federal spending for R & D has remained steady, there has been a marked shift toward support for applied R & D," the report says.
It also stresses the need for new emphasis on software research, saying that the nation has become too dependent on outmoded software, like systems long used by the Federal Aviation Administration and the Internal Revenue Service.
"The FAA and IRS systems have proved to be amazingly difficult to upgrade," the authors write, maintaining that while there have been dramatic improvements in computer hardware in the last 40 years, the ability to develop software has not kept pace.
The report also warns that federal funds for supercomputing have largely collapsed since the breakup of the Soviet Union. The authors argue that failing to support high-performance computing is a national security risk that the nation cannot afford to take.
In order to keep the United States leading in building the world's fastest computers, the nation should set a goal of creating machines capable of reaching the speed of a petaflop -- one thousand trillion mathematical operations a second -- by the year 2010, the report says.
Sites & Affiliations | Leadership | Research & Applications | Major Accomplishments | FAQ | Search | Knowledge & Technology Transfer | Calendar of Events | Education & Outreach | Media Resources | Technical Reports & Publications | Parallel Computing Research Quarterly Newsletter | News Archives | Contact Information