TABLE B.1 Transfers and Other Impacts Since 2003 as Indicated by the New Arrows Added to Figure 1


Research Track and Events

      Origins

      Impact


Digital Communications

Multiple-input and multiple-output

Closely related university research pre-1996; industry R&D (Bell Labs BLAST) in 1998

A fundamental enabler in today’s wireless communications, including WiFi, WiMax, 4G, LTE, and others

Radio-frequency complementary metal-oxide semiconductor integrated circuits

Basic research and some serious engineering efforts in universities throughout the 1990s

The foundational device technology for WiFi, GPS, Bluetooth, and others

Wireless locating

Industry R&D in the early 2000s, with subsequent university research

Fundamental to some wireless systems, and an energy-saving supplement to GPS for mobile devices; now a foundation for growing location-based services

Bluetooth

Ericsson and open SIG 1994, with participation by industry and university; based on earlier research in spread-spectrum radio

Now the industry standard for very-short-range secure networking; used primarily for local connectivity of devices

MACA channel access protocol

University and industry research starting in 1990 and continuing to 2000s

The foundation for collision avoidance on shared physical network links, particularly for wireless networks

Computer Architecture

Adaptive branch predictors

Two-level adaptive branch predictors started in academic research around 1992

Significant influence on subsequent research, as well as influence on modern processor designs

Memory-dependence predictors

Both university and industrial research, emerging in practical form around 1996-1997

Fundamental to enabling aggressive out-of-order execution in practice, affecting research in compilers and architectures, and commercial products

Multicore

Early university-based developments, such as Stanford Hydra and MIT RAW, circa 2002; later, industry R&D such as Sun’s Niagara, around 2005

Fundamental to all processor architectures on the market today; furthermore, the problems of parallel computing have become central in research

Stream-based image-processing architectures

University-based research on stream-processing for image and signal processing, leading to developments such as Imagine, circa 2002

The foundation for today’s GPUs; today there is considerable research and commercial development on applications that go far beyond graphics processing

Graphical processing unit in iPhone

Industry 2007

Present in mobile computers for some time, now considered de rigueur, given the user-experience advantages demonstrated by the iPad/iPhone

Advances in parallel computing applied to commercial product development

Continued university work on parallel computing in 2000s and earlier

Collaboration by top computer architecture researchers in academia and movement between the university and industry communities



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement