Clinton proposes modest increase for HPCC

By Juan Antonio Osuna
CRA Staff

With Clinton and the new Congress scrambling for ways to cut federal discretionary spending, the High-Performance Computing and Communications (HPCC) program received an expectedly modest increase in the president's fiscal 1996 budget.

Released February 6, the Clinton budget proposed a 6% increase for HPCC, which would bring the program to $1.14 billion, up from $1.08 billion in 1995.

Given the tight fiscal environment, HPCC fared better than many other federal science programs. The entire federal budget provides a 3.5% increase for all basic research, just enough to keep up with inflation. But applied research increased 0.8%, disappointing some Republican leaders who favored substantial cuts. Rep. Robert S. Walker (R-PA), chair of the House Science Committee, said, "I regret the strong emphasis [Clinton] places on applied science subsidies."

Overall, academic R&D increased by only 1.1%. (The administration's original budget document erroneously listed a 7% increase.)

The 6% HPCC increase translated into similar increases for the National Science Foundation's Computer and Information Science and Engineering (CISE) Directorate and the Advanced Research Projects Agency's Computer Systems and Communications Technology program.

NSF received an overall increase of 3%, with its research and related activities component receiving a more substantial 7.6%. At a press conference, NSF Director Neal Lane referred to the increase as "good news in tight times."

Under research and related activities, all the directorates received similar increases ranging from 6.7% to 8.5%. Unlike previous years, in which CISE fared better than most directorates, CISE received the smallest increase--6.7%. Its budget increased from $258 million to $276 million.

At the conference, Lane said the slightly different increases among the directorates primarily were a result of constraints imposed by cross-agency initiatives such as HPCC.

HPCC funds within NSF only increased 5.6%. As HPCC represents the bulk of the CISE budget, the HPCC increase was strongly related to the CISE increase. (HPCC does not receive funds in addition to those allocated to various federal agencies; it is simply a way of putting one label on funds spread throughout the government.)

Paul Young, who heads CISE, said the difference in directorate increases "were simply too small to place any [policy] meaning on them."

CISE plans to keep the number of people awarded grants constant but increase the size of these grants enough to cover inflation. The CISE Directorate provides 56% of all federal support to computer scientists at academic institutions. According to CISE estimates, the number of people receiving grants will remain at 5,906, whereas the median grant size will rise from $58,000 in 1995 to $61,000 in 1996.

"Given the relatively modest percentage increases in our research budgets, we expect the average size of grants to go up modestly, partly to match inflation," Young said. "If that holds, it will allow little increase in the number of awardees."

While most components within CISE will sustain growth similar to the overall directorate, a few areas stand out. For instance, funding to support the NSFnet architecture and user connections will increase only 2.2%, whereas basic research in networking and communications will increase 18.7%. The networking and communications research area will increase from $11.24 million to $13.34 million--the largest increase of any subcomponent within CISE.

Young said differences in emphasis among the CISE subcomponents reflect input from various sources, including a meeting at the Airlie Conference Center and the R&D for the NII: Technical Challenges and Realizing the Information Infrastructure reports. The stagnant funding for NSFnet is due to the increasing privatization of network infrastructure, he said. "As that program winds down, you'll see more emphasis on experimental networks" and basic research.

"In general, we have continued the trend toward increased emphasis on research for information infrastructure," Young said. "NSF funds a major portion of all of the nation's basic research in networking and communications. These are clearly going to play a continuing major role, as today's networks are transformed to support the information infrastructure technologies in the 21st century."