Computing and the National Science Foundation, 1950-2016. William Aspray

Чтение книги онлайн.

СКАЧАТЬ CISE should be governed by considerations of their fit to the respective missions of the directorates, and of their relative contributions to strengthening U.S. research in computing or in engineering,”¹⁶⁷ suggested program director Howard Moraff. He also raised concerns about possible disruptions to the new programs created in Nam Suh’s recent reorganization of Engineering and urged collaboration between CISE and Engineering.

      Bell wrote to Bloch defending his approach to organizing the directorate, suggesting:

      The rationale for Engineering disciplines in CISE is that contemporary engineering naturally divides into two parts: those research areas based on the physical and biological sciences and those which deal with information. This provides a working boundary: on one side are research areas governed by the physical transformations, and on the other side those research areas concerned with the transformation of information from one form to another.¹⁶⁸

      In Bell’s personal notes dated March 20, 1986, he characterized the Nam Suh position as being that CISE should be “pure computer,” while Bell himself thought it should be “pure information processing (and storage, transmission, switching, transduction), robotics, and some cross-disciplinary programs that make significant use of the computer.” In these notes, Bell discussed other organizations of traditional EE and computing, and the National Academy of Engineering taxonomies.

      On March 24, Frank Huband responded to Nam Suh¹⁶⁹ concerning the probable decision by Bloch to transfer Communication Systems and Signal Processing programs (CSSP) to CISE. Huband was concerned that “important parts of the current CSSP program will not be relevant to the purposes of CISE and may thus not be eligible for future funding.” This memorandum was forwarded to Bloch and Bell, and Bell forwarded it to Chuck Brownstein, Bernie Chern (who had been reassigned from the Computer Engineering program to CISE), and me. Bell commented that “we will work it out over the next couple of months along with MOSIS funding, etc.”

      Bell sent his initial plan¹⁷⁰ for CISE to Bloch for approval via the Assistant Director for Administration on April 17, 1986. This memorandum proposed transferring the Division of Computer Research (DCR) from Mathematical and Physical Sciences, the Division of Information Science and Technology (DIST) from Biological and Behavioral Sciences, and the Office of Advanced Scientific Computing (OASC) from the Director’s Office intact. It proposed creating a new Division of Computer and Information Engineering (CIE) to temporarily house the computer engineering programs (Software Systems Design; Computer Systems Architecture; Vision, Robotic, and Knowledge-based Systems) and the Communications and Signal Processing programs, each from the Engineering Directorate. The proposal added a Division Director (Bernard Chern) for the new CIE Division, a CISE Planning Officer (Jerry Daen), an acting CISE Executive Officer (Charles Brownstein, who remained Director/DIST), and an acting senior scientist for planning and program development (W. Richards Adrion, who remained Deputy Director/DCR). The new directorate would have 54 positions, 49 permanent and 5 IPAs. Bloch approved, and the directorate was officially launched on May 1, 1986. At the time it was officially created, CISE had two advisory committees: Computer Research (mainly associated with DCR) and Advanced Scientific Computing (mainly associated with OASC).

      In Bell’s presentation¹⁷¹ to the National Science Board in May, he described using the CISE research budget as a “balance wheel” to DARPA and industry. He defined five research areas: parallelism as applied to parallel processing; automation, robotics, and intelligent systems; ultra-large-scale integrated systems; advanced scientific and engineering computing; and networks and distributed computing. Bell focused on these areas because they had relatively clear, long-term goals; measurable output; an emphasis on maintaining U.S. leadership in computing; significant economic and competitive impact; and a demand for undergraduate and graduate training. Across these five initiatives, CISE would support basic, front-end research throughout the entire computer research community at a time when DARPA was becoming more “mission oriented.”

      Between April and August, a number of organizational structures were considered. Bell’s five initiatives helped structure the divisional organization of CISE. While the initiatives were cross-cutting, divisions were thought of as the leads: DCR for parallelism as applied to parallel processing; DIST for automation, robotics, and intelligent systems; CIE for ultra-large-scale integrated systems; and OASC for advanced scientific and engineering computing. Bell did not see OASC leading networks and distributed computing, and that eventually led to a fifth division in CISE. DIST leadership in automation, robotics, and intelligent systems led to a proposed ARIS division that eventually became Information, Robotics, and Intelligent Systems (IRIS) due to a continuing commitment to information sciences and systems. Initially CIE was to become the Ultra-Large-Scale Integration (ULSIS) Division; but with a responsibility for computing design and architecture, it was renamed the Microelectronic Information Processing Systems (MIPS) Division. The need to locate the communication and signal processing programs led to DCR taking on that responsibility as the Computer and Communications Research (CCR) Division. One other proposed restructuring would have divided OASC¹⁷² into three sections: Centers with Larry Lee as Head; Networking and Distributed Computing with Steve Wolff as Head: and a New Technologies Section with Al Harvey as Head. The plan included research and EXPRES program directors. The tension over NSFNET as a national vs. supercomputer network continued and John Connolly strongly objected.

      On August 26, 1986, Bell proposed¹⁷³ restructuring CISE by reconfiguring the DCR, moving the Intelligent Systems program to DIST, renaming DCR as the Division of Computer and Computation Research, restructuring DIST and renaming it the Division of Information, Robotics, and Intelligent Systems, and restructuring CIE and naming it the Division of Microelectronic Information Processing Systems. The Office of Advanced Scientific Computing was renamed the Division of Advanced Scientific Computing. In the divisions, programs were restructured to reflect a new divisional mission. This plan was approved and became official¹⁷⁴ on October 17, 1986.

      This reorganization resulted in some personnel changes. Earlier in July, Chuck Brownstein officially was named CISE Executive Officer. (He had served briefly as Acting AD/CISE until Bell was sworn in on June 17, 1986.) While I would resign officially on September 14,¹⁷⁵ remaining on a part-time basis through the fall, Gordon Bell already knew this when he wrote his August 26 memorandum. He officially transferred me into the position from DCR and argued to keep the position after my departure. Kent Curtis filled this position a year later. EXPRES was moved from DASC into the AD’s office. The last organizational change occurred on December 10, 1986, when the Networking and Communications programs in DASC became a fifth division in CISE: the Networking and Communications Research and Infrastructure Division (NCRI), with Steve Wolff as DD.

      While the organizational debates were going on, the staff that were clearly moving to CISE began to address the planning process outlined by Bridgewater. I wrote¹⁷⁶ in March 1986 to Chuck Brownstein, Bernie Chern, John Connolly, and Kent Curtis about long range planning for CISE, asking them to produce planning documents covering both existing programs and potential new initiatives. This was a three-part request: an exercise to “define the base,” long-range strategic planning, and issue papers on new initiatives. All of these were due March 17.

      I wrote¹⁷⁷ again to the (unofficial) CISE division directors about the need for them to develop plans to address the five Bell initiatives, asking for two-page position papers. Each acting division director needed to answer four questions Bloch had asked each directorate to address: What difference has NSF support made? What is the NSF role in [discipline] research? What are the programmatic gaps? What are your priorities in the event of reductions? Table 2.2 includes excerpts from the Long-Range Planning Material submitted to the Office of Budget, Audit, and Control¹⁷⁸ in April 1986.

      Each of five research initiatives included research opportunities/breakthroughs needed, current efforts, and plans and initiatives. Two-page position papers were written on (1) parallelism (lead: Rick Adrion); (2) advanced scientific computing СКАЧАТЬ