Dr. Dobb's Journal January 2002
A coalition of companies including Philips, Intel, and Microsoft have announced their intention to develop a new technique for adding speech recognition and telephony capabilities to web applications. The proposed Speech Application Language Tags (SALT) would extend existing markup languages such as HTML and XML, and is intended to function as a lightweight alternative to VoiceXML. Microsoft has stated that it plans to add SALT support to Visual Studio.NET.
Forum members see SALT as a way to solve the input problem for wireless devices; to make speech capabilities available to a wider range of developers; and to integrate the Web more tightly with telephone networks. "End users will be able to use SALT-based applications any time, anywhere, and from any device using speech, text, or graphical interfaces independently or at the same time" promises the SALT Forum.
The new markup will be royalty free and, since it is based on HTML, platform independent. The SALT specification is expected to be completed by the first quarter of 2002, and the forum plans to submit it to "a standards body" by midyear. For more information, see http://www.saltforum.org/.
The Computer Museum History Center bestowed three Fellow Awards this year, honoring Frederick Brooks, Maurice Wilkes, and Jean Sammet. Museum Fellows are chosen for fundamental, long-range contributions to the computer field. At least 10 years must elapse between the individual's achievements and the bestowal of the award.
Frederick Brooks is well known for his research into the practice of software engineering. He is the author of The Mythical Man-Month and other books, and he was awarded the Turing Award in 1999. Jean Sammet, who served as a member of the committee that developed Cobol, conceived and directed the development of FORMAC, an early language for symbolic mathematics. She also led the IBM Federal Systems Division work on Ada, and wrote a classic 1969 book on programming languages.
Maurice Wilkes worked on EDSAC, a computer which first became functional in 1949. He is quoted in museum literature as saying, "It was on one of my journeys between the EDSAC room and the punching equipment that...the realization came over me with full force that a good part of the remainder of my life was going to be spent finding errors in my own programs." For more information, see http://www.computerhistory.org/.
The Terascale Computing System (TCS) has been successfully installed at the Pittsburgh Supercomputing Center (http://www.psc.edu/). The proud owners call the new supercomputer "the most powerful system in the world committed to unclassified research." With 6 teraflops of processing power, the TCS is surpassed only by Lawrence Livermore National Laboratory's ASCI White supercomputer. ASCI White, however, is devoted to classified weapons research, including the simulation of nuclear explosions. The TCS is available on a grant basis to support public research. Projects supported by the TCS include earthquake modeling, storm-scale weather forecasting, global climate change, and protein genomics modeling.
The TCS is constructed of 3000 Compaq Alpha EV68 microprocessors (with peak floating-point capability of 2 gigaflops) housed in 750 four-processor AlphaServer systems running Tru64 UNIX. The switch structure was provided by Quadrics Supercomputers World (http://www.quadrics.com/). The TCS has 3 terabytes of memory and can write the entire memory to disk in under 40 seconds. It takes up the floor space of a basketball court, uses 21 miles of cable, requires 664 kilowatts of power, and produces heat equivalent to burning 169 pounds of coal an hour.
The supercomputer was designed and constructed by Compaq and the Pittsburgh Supercomputing Center, with funding from the National Science Foundation. It cost $45 million to build. Grants for research time on the machine are available through the National Science Foundation's Partnerships for Advanced Computational Infrastructure (http://www.interact.nsf.gov/cise/descriptions.nsf/pd/paci/).
A new collection of essays about the WWII codebreaking effort entitled Action This Day: Bletchley Park From the Breaking of the Enigma Code to the Birth of the Modern Computer (Bantam Press, 2001; ISBN 0593049101) suggests that the Allies would have understood the German Enigma machine much earlier if they had listened to an unknown woman.
According to historian Ralph Erskine (who edited the book along with Michael Smith), Dilly Knox and Alan Turing were stumped by the wiring of the machine's keyboard to its rotors. It was not until July of 1939 that Knox learned from Polish codebreakers that the keys were actually wired to the rotors in alphabetical order.
A news story in The Guardian quotes Erskine as writing, "Surprisingly, one cryptanalyst at GC&CS, a 'Mrs. BB' (it has been impossible to identify her) 'had seriously contemplated' that the wiring was indeed an identity [A to A etc.]. But she had not been given a crib [the plain text of earlier German messages encyphered with the help of the Poles]...either because organisation was not Knox's forte, or because he thought that she would be wasting her time in following it up. She had, therefore, been unable to make any progress."
Scientists from Bell Labs have created organic transistors with a single-molecule channel length. The size of a transistor's channel (that is, the space between its electrodes) influences its output current and switching speed.
Using the tiny transistors, which are approximately a million times smaller than a grain of sand, Bell Labs scientists Hendrik Schon, Zhenan Bao, and Hong Meng built a voltage inverter, a standard electronic circuit module commonly used in processors that converts 0s and 1s. Though just a prototype, the success of the simple circuit suggests that molecular-scale transistors could one day be used in microprocessors and memory chips, squeezing thousands of times as many transistors onto each chip than possible today.
The main challenges faced in making molecular-scale transistors are fabricating electrodes separated by only a few molecules and attaching electrical contacts to the tiny devices. The Bell Labs researchers were able to overcome these hurdles by using a self-assembly technique and a design in which each electrode is shared by many transistors. For more information, see http://www.bell-labs.com/news/2001/october/17/1.html.