William Kahan

Article on other languages:

• William_Kahan
• William_Kahan
• ウィリアム・カハン
• William_Kahan
• William_Kahan
• Кэхэн,_Уильям_Мортон
• Вилијам_Кон

	del.icio.us
	Digg
	Furl
	Reddit
	Rojo
	Add to OnlyWire

William Morton Kahan
Born	June 5, 1933 (1933-06-05) (age 75) Toronto, Ontario
Nationality	Canadian
Fields	Mathematics Computer Science
Institutions	University of California, Berkeley
Alma mater	University of Toronto
Known for	IEEE 754 Kahan summation algorithm
Notable awards	Turing Award ACM Fellow

William Morton Kahan (born June 5, 1933, in Toronto, Ontario, Canada) is a mathematician and computer scientist whose main area of contribution has been numerical analysis. Among his colleagues he is known as Velvel Kahan.

He attended the University of Toronto, where he received his Bachelor's degree in 1954, his Master's degree in 1956, and his Ph.D. in 1958, all in the field of mathematics.

Among his many contributions, Kahan was the primary architect behind the IEEE 754-1985 standard for floating-point computation (and its radix-independent follow-on, IEEE 854) and developed the Kahan summation algorithm, an important algorithm for minimizing error introduced when adding a sequence of finite precision floating point numbers.

In the 1980s he developed the program "paranoia", a benchmark that tests for a wide range of potential floating point bugs.

He received the Turing Award in 1989, and was named an ACM Fellow in 1994.

Kahan is now a professor of mathematics, computer science, and electrical engineering at the University of California, Berkeley, and continued his contributions to the IEEE 754 revision that led to the current IEEE 754 standard. He has been called “The Father of Floating Point,” since he was instrumental in creating the original IEEE 754 specification.

He is an outspoken advocate of better education of the general computing population about floating-point issues, and regularly denounces decisions in the design of computers and programming languages that may impair good floating-point computations.

He coined the term "The Table-Maker's Dilemma" for the unknown cost of rounding transcendental functions:

"Nobody knows how much it would cost to compute y^w correctly rounded for every two floating-point arguments at which it does not over/underflow. Instead, reputable math libraries compute elementary transcendental functions mostly within slightly more than half an ulp and almost always well within one ulp. Why can't Y^W be rounded within half an ulp like SQRT? Because nobody knows how much computation it would cost... No general way exists to predict how many extra digits will have to be carried to compute a transcendental expression and round it correctly to some preassigned number of digits. Even the fact (if true) that a finite number of extra digits will ultimately suffice may be a deep theorem."^[1]

When Hewlett-Packard introduced the original HP-35 pocket scientific calculator, its numerical accuracy, in evaluating transcendental functions, for some arguments, was not optimal. Hewlett-Packard worked extensively with Kahan to enhance the accuracy of the algorithms, which led to major improvements. This was documented at the time in the Hewlett-Packard Journal^[2][3].

References

External links

• William Kahan's home page
• portrait picture of Kahan
• Paranoia for modern graphics processing units (GPUs)
• Paranoia source code in multiple languages

See

• Richard Karpinski. 1985. Paranoia: A floating-point benchmark. Byte Magazine 10, 2 (Feb.), 223-235
• IEEE. 1987. IEEE standard for binary floating-point arithmetic [url]. ACM SIGPLAN Notices 22, 2 (Feb.), 9-25
• An Interview with the Old Man of Floating-Point, 1998-Feb-20 [1]
• A Conversation with William Kahan, Dr. Dobb's Journal November, 1997