Charles R. Severance

FireWire finally comes home

The article discusses the IEEE 1394 networking standard, otherwise known as FireWire. FireWire was designed to link personal computers, digital cameras, televisions, DVD players, printers, and other home electronics equipment. At the time FireWire seemed like a perfect idea whose time had come, but there was one small problem: a competing technology, called Universal Serial Bus (USB), promised to do nearly the same thing. Like FireWire, USB can connect multiple peripherals to a single port on the back of our computers. But USB is designed to be a simpler, slower interface that is less expensive to manufacture. Because both technologies arrived at about the same time, hardware and software vendors had to make a choice as to which technology to develop and support. USB got the nod in the Wintel world-with Windows driver support for USB peripherals as early as October 1996-but it wasnt until Windows 98 that consumers got full OS support for USB. Early support on the Wintel platform, however, could account for the large number of USB components that have been popping up over the past two years, such as speakers, joysticks, printers, video cameras, and the like, all of which have made it to market much more quickly than FireWire based technologies. The prospect of the increasing use of FireWire is discussed.

Managing the Overall Balance of Operating System Threads on a Multiprocessor Using Automatic Self-Allocating Threads (ASAT)

Threads working on compute intensive tasks work best when there is one thread performing real work on each processor. The challenge is to schedule threads to maintain one running thread per processor by dynamically adjusting the number of threads as the overall load on the machine changes. This work identifies the situations on a multiprocessing system when the operation of a lightweight thread compiler run-time environment might be negatively impacted by other threads running on the system. A solution called Automatic Self-Allocating Threads (ASAT) is proposed as a way to balance the number of active threads across a multiprocessing system. Our approach is significant in that it is designed for a system running multiple jobs, and it considers the load of all running jobs in its thread allocation. Our approach uses self-scheduling, so it is implemented in a run-time environment rather than in an operating system. Finally, the self-scheduling means that jobs need not all be scheduled by ASAT.

Could LDAP be the next killer DAP

About a year ago, Netscape Communications, with the support of more than 40 other companies, adopted the Lightweight Directory Access Protocol (LDAP). Notably absent from the list of early LDAP supporters, Microsoft came on board eight months later. LDAP is a protocol that allows a program such as a browser or an e-mail package to perform directory lookups across a wide variety of directories, even if they run on different operating systems and directory environments. For example, LDAP could search the internal corporate directory stored on a corporate server or a publicly available directory. In some ways, LDAP itself is not the critical element. A protocol alone is not enough. Whats missing is a useful set of directory servers that support LDAP and client software that conveniently queries those servers using LDAP. Well, as it turns out both of those things do exist. In Netscape Navigator 4.0, simply go to Edit I Search Directory, select the proper directory service, type in a persons name, press Search and voila-you get a short list of the people who might match the name in question. From the interface, you are now two clicks away from sending an e-mail message to that person. Now that is a killer DAP-its clearly obvious just how useful this feature will be.

Linking computers and consumer electronics

The capabilities of our personal computers have increased dramatically over the past 15 years, and so has the number of connectors on the back of our systems. Originally, we needed a serial port for a modem and a parallel port for a printer. Now we also need ports for a mouse, audio input, audio output, video input, video output, Ethernet, and a camera. Enter the IEEE 1394 standard known as Firewire. This standard proposes to provide a single port on the back of our computers that can handle nearly all of the communication for which we now need eight to 10 ports. Along with connecting standard peripherals to computers, Firewire is designed for use with consumer audio, video, and television equipment.

IEEE 754: An Interview with William Kahan

Standards I f you were a programmer using floating point computations in the 1960s and 1970s, you had to cope with a wide variety of configurations, with each computer supporting a different range and accuracy for floating-point numbers. While most of these differences were merely annoying, some were very serious. One computer, for example, might have values that behaved as non-zero for additions but behaved as zero for division. Sometimes a programmer had to multiply all values by 1.0 or execute a statement such as X = (X + X) − X to make a program work reliably. These factors made it extremely difficult to write portable and reliable numerical computations. In 1976, Intel began to plan for a floating point coprocessor for the Intel i8086/8 and i432 microprocessors. John Palmer convinced Intel that they needed to develop a thorough standard to specify the arithmetic operations for their coprocessor so that all Intel processors would produce the same results. Because William Kahan had extensive experience with the IBM, Cray, and Control Data Corp. (CDC) floating point, he was one of the few who understood the challenges of writing accurate numerical code. In 1976, Kahans influence on floating-point processing escalated when Intel hired him as a consultant to help design the arithmetic for the 8087 processor. As a result, he had a hand in the birth of the IEEE 754 specification for floating point computations. —Charles Severance Charles Severance: When Intel hired you as a consultant in 1976, what did they want you to do? William Kahan: The folks at Intel decided that they wanted really good arithmetic. The DEC VAX was really not that bad, so my reasoning went: Why not copy the VAX? Intel wanted the best arithmetic, so Palmer and I got together to think about what the best arithmetic should be. One of the things Palmer told me was that Intel anticipated selling these coprocessors in very large numbers. The best arithmetic was what was best for a large market, which subsequently started to frighten Silicon Valley because of rumors that Intel was building floating point on a single chip, the i8087. And when they heard rumors of what was going to be on that chip, they were aghast. CS: Out of this thinking grew IEEE 754? WK: People have said from time to time (as a joke) that the other Silicon Valley companies got worried and …

Web-based education experiences

The World Wide Web has the potential to revolutionize instruction and increase educational opportunities across corporate, government and educational sectors. Not only will the Web enable distance learning and peer-to-peer connectivity, but it will also enhance student feedback and let instructors understand whether and how long students have studied a particular lesson. It also has the potential to reduce training costs. The authors discuss the creation of an online learning environment considering practical and pedagogical issues.

Comparing gang scheduling with dynamic space sharing on symmetric multiprocessors using automatic self-allocating threads (ASAT)

The work considers the best way to handle a diverse mix of multi-threaded and single-threaded jobs running on a single symmetric parallel processing system. The traditional approaches to this problem are free scheduling, gang scheduling, or space sharing. The paper examines a less common technique called dynamic space sharing. One approach to dynamic space sharing, automatic self allocating threads (ASAT), is compared to all of the traditional approaches to scheduling a mixed load of jobs. Performance results for ASAT scheduling, gang scheduling, and free scheduling are presented. ASAT scheduling is shown to be the superior approach to mixing multi-threaded work with single threaded work.

The value of the formal standards process

The purpose of this article is to critique the process of developing formal standards, which are those that have been approved by an official standards-making body. The bodies that impact US computer standards most include the International Standards Organization (ISO), the International Telecommunications Union (ITU), the American National Standards Institute (ANSI) and the Institute of Electrical and Electronics Engineers (IEEE). Other groups develop important standards that are outside the formal process. These include the Open Software Foundation (OSF), X/Open, and the Internet Engineering Task Force (IETF). The main difference between a formal standards organization and the other groups is the legal framework in which the body operates. The formal organizations are often chartered by the government with strict procedures and rules imposed on the standards development process. OSF and X/Open are each directed by a board of directors, whereas the IETF is an independent, self-governing body that develops its own rules and procedures. >

Posix: a model for future computing

The author considers how the Posix reference model remains accurate while other standards are becoming obsolete. The reference model identifies four main interfaces at which standards really matter: application programming interface; user interface; data interface; and communications interface.

Automatic self-allocating threads (ASAT) on an SGI Challenge

Automatic self allocating threads (ASAT) is proposed as a way to balance the number of active threads across a shared-memory multiprocessing system. Our approach is significant in that it is designed for a system running multiple jobs, and it considers the load of all running jobs in its thread allocation. In addition, the overhead of ASAT is sufficiently small so that the run times of all jobs improve when it is in use. We consider the application of ASAT for improving the scheduling of threads on an SGI Challenge. We demonstrate how the number of threads of an ASAT job adjusts to the overall system load to maintain thread balance and improve system throughput.