Jump to content

Peter Kogge

From Wikipedia, the free encyclopedia

This is an old revision of this page, as edited by KasparBot (talk | contribs) at 03:42, 15 May 2016 (migrating Persondata to Wikidata, please help, see challenges for this article). The present address (URL) is a permanent link to this revision, which may differ significantly from the current revision.

Peter Kogge
EducationComputer Science and Engineering University of Notre Dame, PhD Computer Science Stanford University
OccupationComputer engineer

Dr. Peter Michael Kogge is an American computer engineer and IBM Fellow.

Background

Dr. Kogge has been at the forefront of several innovations that have shaped the computing industry over the past three decades. While working on his PhD at Stanford in the 1970s, Dr. Kogge invented what is still today considered the fastest way of adding numbers in a computer, the Kogge-Stone Adder process, an approach still used in microprocessors by Intel and other companies and generally considered the fastest adding design possible.

After receiving his degree, Dr. Kogge joined the computer engineering team at IBM. During his time there, he was a co-inventor on over three dozen patents. His design of the Space Shuttle I/O processor at IBM was one of the first multithreaded computers, and the first to fly in space.

Contributions

Peter was the author of the first textbook on pipelining, a now ubiquitous technique for executing multiple instructions in a computer in parallel. At IBM, Kogge was also the inventor of the world's first multi-core processor, EXECUBE, which Kogge and his team placed on a memory chip in an early effort to solve the data bottleneck problem that Emu is solving today.

In 1994, Dr. Kogge joined the University of Notre Dame as a faculty member, the Ted H. McCourtney Professor of Computer Science and Engineering.[1] Most recently, he led a team of some the nation’s most renowned computer professionals, commissioned by the U.S. Defense Advanced Research Projects Agency (DARPA), to explore what sorts of technologies engineers would need to build a supercomputer capable of executing a quintillion (a billion billion) mathematical operations per second. The results of this study were a cover story on the Institute for Electrical and Electronic Engineers (IEEE) Spectrum magazine in February, 2011.

References

  1. ^ "Notre Dame News". nd.edu. University of Notre Dame.