The Impact of Academic Research on Industrial Performance (2003) / Chapter Skim
Currently Skimming:
2. Report of the Panel on the Network Systems and Communications Industry
2. Report of the Panel on the Network Systems and Communications Industry
Pages 29-76
The Chapter Skim interface presents what we've algorithmically identified as the most significant single chunk of text within every page in the chapter.
Select key terms on the right to highlight them within pages of the chapter.
From page 29... ...
The questionnaire was followed by a workshop attended by approximately 30 senior individuals in the network systems and communications sector (see Addendum)
|
From page 30... ...
It clearly includes the manufacturing of telecommunications equipment and the services that use such equipment, such as telephony, wireless telephony, broadcast television, cable and satellite television, radio, and Internet service. Both the equipment and services sectors increasingly require computing equipment and software, and, in fact, the computer and communications industries are no longer separate industries.
|
From page 31... ...
These three trends convergence, embedding, and network applicationscharacterize the network systems and communications sector. The panel's assessment of the contributions of academic research to this industry is based on this broad definition.
|
From page 32... ...
Size Because our definition has vague boundaries and because the industrial classifications used to gather statistics have not been adapted to the rapid changes in the industry, it is difficult to determine the size of the network systems and communications sector. Table 2-1 summarizes sales and employment in the information technology industry based on Bureau of the Census data (U.S.
|
From page 33... ...
Custom integrated circuits can implement very complex communications functions; coupled with custom-built and proprietary software designs, equipment vendors compete intensely in terms of technology, reliability, and cost of ownership. Another important feature of the network systems and communications sector is its reliance on components with well defined interfaces.
|
From page 34... ...
In some cases, networking has focused on providing Internet access for employees and customers; in others, the focus has been on the development of internal networks linking production and distribution facilities across the company. So far, neither academic nor industrial research has addressed the problems of service delivery in a structured and sustained manner.
|
From page 35... ...
Research results may be the basis for some of these improvements, and research has achieved major breakthroughs in these areas; this research is performed or funded by materials, equipment, and microchip fabricators, not by the telecommunications equipment manufacturers (see Box 2-1~.
|
From page 36... ...
Dell Computer, for example, does not have in-house R&D; in effect, Dell is a broker that negotiates attractive deals to buy components and computer-assembly services for its build-a-computer-to-order business. Dell depends on R&D investments by its vendors, especially Intel and Microsoft, that make the microprocessors and operating system software on which the personal computer business depends.
|
From page 37... ...
Industrial research is concentrated in the laboratories of a few of the largest companies, such as Intel, Microsoft, IBM, Compaq, Lucent, AT&T, HewlettPackard, Sun Microsystems, and Xerox. Although many of these firms invest 10 to 15 percent or more of revenues in R&D each year, the vast majority of this is for "development," that is, for the engineering of the next generation of products.
|
From page 38... ...
, Trek (Rochester) Atari, Nintendo, SGI, Pixar ARPANET, Aloha, Internet Pup DECnet, TCP/IP Rings, Hubnet Ethernet, Datakit, Autonet LANs, switched Ethernet Lisp machine, Stanford Xerox Alto ( to World Wide Web 1980 1985 1990 1995 2005 Xerox Star, Apollo, Sun Engelbart / Rochester Alto, Smalltalk Star, Mac, Microsoft Berkeley, Caltech, MOSIS many Berkeley, Stanford IBM 801 SUN, SGI, IBM, HP FIGURE 2-1 Examples of academic government-sponsored (and some industrysponsored)
|
From page 39... ...
. , Speech recognition Broadband in last mile 1965 1970 1975 \\ · ~ ~ ~ ~ 1980 1985 1990 1995 2005 Berkeley, Wisconsin IBM Oracle, IBM, Sybase Tokyo, Wisconsin, UCLA IBM, ICE JCL, Teradata, Tandem Wisconsin, Stanford IBM, Arbor IRI, Arbor, Plato Illiac 4, CMU, Caltech, HPC IBM, Intel CM-5, Teradata, C ray T3D Berkeley Striping/Datamesh, Petal many Berkeley, Purdue (COMA)
|
From page 40... ...
Some of the features of that culture are described below. A 1995 report of the National Research Council Computer Science and Telecommunications Board documents the effect of the research culture in the computer and high-performance computing arenas (NRC, 1995~.
|
From page 41... ...
Market explodes. Industrial research the gaps [lots of difficult research here]
|
From page 42... ...
Anecdotal evidence also suggests that many industry researchers have found places in universities as industrial research spending has declined at telecommunications and computer hardware companies. Their knowledge and industrial experience can provide valuable insight for academic research.
|
From page 43... ...
The success of the deregulation of telecommunications services depends on standard interfaces. For example, a compatible local exchange carrier must be able to connect into the networks of other local exchanges and long-distance carriers with predictable interfaces.
|
From page 44... ...
helped equip academic research centers with the new technology, which served as a substrate for academic research in networking and interactive computing. Subsequently, parallel computers were provided to encourage research on software tools for writing highperformance parallel computing applications.
|
From page 45... ...
Although the volume of R&D investment in computer-related industries has kept pace with the growth of business over the past decade, the R&D spending of the telecommunications component of the network systems and communications sector has contracted in the wake of AT&T's divestiture, deregulation, and most recently, deep recession in the telecommunications industry. Although the amount of industry support for university research in network systems and communications is not known, overall industry support for research
|
From page 46... ...
NSF supports university-based engineering research centers and science and technology centers that must have industry contributions to supplement government funding. Initiatives like Internet2 and Next Generation Internet, which are funded principally by government, solicit industry support.
|
From page 47... ...
the network systems and communications sector has benefited greatly from a national research culture in which individuals move frequently between academia and industry, thereby increasing their knowledge of both and their contributions to both; (2) personal relationships are crucial; and (3)
|
From page 48... ...
Graduates with advanced degrees have already shown greater than average ability; and research training is considered evidence that an individual can tackle difficult technical problems, such as designing and building complex systems. Students of electrical engineering and computer science have typically been in great demand, not only by companies in the network systems and communications sector, but also by other companies trying to modernize their information technology.
|
From page 49... ...
Most of the research was necessarily conducted in academia, because industry typically does not invest in risky research that offers only long-term prospects for payoff. The two case studies below illustrate how academic research has contributed to the network systems and communications sector: (1)
|
From page 50... ...
and released in the late 1970s, it was easy to connect academic research computers to the network. Ad hoc committees of academic researchers refined the TCP/IP protocol standards, including application protocols.
|
From page 51... ...
DARPA and NSF, with university and industry support, organized a series of test beds to explore high-speed networking technologies and test emerging products and protocols. Between 1990 and 1994, NSF and DARPA funded the Gigabit Testbed Initiative, a university-industrygovernment effort to explore networking technologies at speeds of 155 Mb/s and higher; one of these test beds achieved long-haul transmission at 800 Mb/s.
|
From page 52... ...
Today, these committees have broad participation from academia, industry, and nonprofit organizations. Academic Contributions An SRI study commissioned to analyze the nature of the research that contributed to the Internet described the contribution of academic research in some detail (SRI International, 1997~: The Internet appears, overall, to be primarily a problem-driven, technology-based innovation that required little direct input from fundamental research for its realization.
|
From page 53... ...
Cioffi used funds from an NSF Presidential Investigator Award (1987-1992) , with matching funds from several companies, including Bell Communications Research, to investigate asymmetric digital subscriber lines.
|
From page 54... ...
In March 1998, Texas Instruments acquired Amati and the DMT license for approximately $450 million in cash. At the end of 2001, 3.6 million residential DSL (digital subscriber lines)
|
From page 55... ...
Management, Design, and Policy Academic research in a variety of nonengineering disciplines has also contributed to the network systems and communications sector. This research has focused on how computers and communications systems fit into larger socioeconomic systems.
|
From page 57... ...
, The HomeNet Field Trial of Residential Internet Services (Kraut et al., 1996) , and The Second Self (Turkle, 1984)
|
From page 58... ...
, showed how studies in cognitive psychology could be used to estimate human performance when interacting with a computer. These and other performance studies have influenced the design of graphical user interfaces.
|
From page 59... ...
The network systems and communications sector does not have an institution comparable to the Semiconductor Research Corporation, an organization that provides industrial support for university research relevant to semiconductors, based on a 10-year technology "road map" to help guide research funding decisions (Bailey et al., 1998~. Although the establishment of a consortium of network systems and communications businesses has been discussed, nothing has come of it so far.
|
From page 60... ...
FINDINGS AND RECOMMENDATIONS Academic research has made essential contributions to the network systems and communications sector. Although these contributions trained researchers, new technologies, algorithms, and prototype systems; early operating experience; studies of social and economic effects cannot be quantified, they have undoubtedly had a substantial impact.
|
From page 61... ...
University-industry collaborations are fostered by a vigorous research culture, and academic research has been crucial to the technical evolution of the industry, especially in the development and deployment of the Internet. To be sure, the recent deep recession in the telecommunications sector, which has forced significant reductions in corporate R&D budgets and manpower, has further diminished the industrial research contributions of this important subsector of the network systems and communications industry a trend begun with the breakup of the Bell system and further deregulation.
|
From page 62... ...
to support small portable devices spawned breakthrough products, such as pagers, cellular telephones, and packetradio modems. Innovations in the network systems and communications industry have been characterized by the integration of a wide range of technologies: chip designs with increasing levels of integration; digital-analog integration in wireless and wireline communications (e.g., cellular and satellite telephony, wireless devices, such as pagers and security devices, modems and cable modems, local-area network and intranetwork systems and communications receivers, optical network interfaces)
|
From page 63... ...
; the ATM Forum, which was organized to promote data-networking uses of Asynchronous Transfer Mode (ATM) ; and the discrete multilane modulation standard for asymmetric digital subscriber lines, have academic participants.
|
From page 64... ...
Finding 2-7. Many Internet service providers are not willing to make their data available to researchers.
|
From page 65... ...
Although data are available for some experimental networks, many Internet service providers have not been willing to make their data available to researchers. This has hampered university research that might lead to improved network design and operation.
|
From page 66... ...
To maintain the pipeline of both academic and industrial researchers, the following measures could be taken: . Universities should Provide early research experiences for undergraduates or even secondary school students.
|
From page 67... ...
Encourage sabbaticals in both directions, enabling academics to spend time in industry, especially in start-up companies. Support people and projects that involve academic and industry researchers in essential ways.
|
From page 68... ...
3The loss of faculty to commercial endeavors was limited in time and to only a few programs. Data from the most recent Taulbee Survey of computer science and computer engineering departments indicate that faculty numbers have grown and are anticipated to grow through 2004.
|
From page 69... ...
1996. The HomeNet field trial of residential Internet services.
|
From page 70... ...
1998. Diagram presented during panel session on Changing the Interaction Between Academic Research and Industry: University, Industry and Government Perspectives at the workshop How Can Academic Research Best Contribute to Network Systems and Communications?
|
From page 71... ...
THE IMPACT OF ACADEMIC RESEARCH ON INDUSTRIAL PERFORMANCE NETWORK SYSTEMS AND COMMUNICATIONS PANEL We invite your responses to the following questions. Your responses will be used by our Panel as background information for our report.
|
From page 72... ...
10. Do you see any downside to enhanced university-industry research collaboration?
|
From page 73... ...
Contributions and impacts of academic research on performance in the network systems and communications industry: Design, Social, Management, and Policy Sciences Dan Atkins, Walter Bender, Robert Kraut, Tom Malone, Marvin Sirbu 1:30 pm Session III. Structures for university-industry collaboration James Flanagan, Stewart Personick, David Roessner, Donald Strickland, Stephen Wolff 2:30 pm Break
|
From page 74... ...
74 THE IMPACT OF ACADEMIC RESEARCH ON INDUSTRIAL PERFORMANCE 2:45 pm Session IV. Changing the interaction between academic research and industry: University, Industry, and Government Perspectives Hamid Ahmadi, Ed Lazowska, James Morris, Rick Rashid, George Strawn, David Tennenhouse 4:30 pm Discussion, conclusions and recommendations Bob Sproull
|
From page 75... ...
Chairman Emeritus Bellcore Robert Kahn Corporation for National Research Initiatives Robert Kraut Department of Social and Decision Sciences Carnegie Mellon University H.T. Kung Department of Electrical Engineering and Computer Science Harvard University Ed Lazowska Department of Computer Science and Engineering University of Washington Tom Malone Sloan School of Management Massachusetts Institute of Technology David Mills Department of Electrical and Computer Engineering University of Delaware
|
From page 76... ...
Strickland Chair, Management Department Southern Illinois University David Tennenhouse Defense Advanced Research Projects Agency Stephen Wolff * Executive Director Advanced Internet Initiatives Division Cisco Systems, Inc.
|
Key Terms
This material may be derived from roughly machine-read images, and so is provided only to facilitate research.
More
information on Chapter Skim is available.