I’m Philip Emeagwali. I invented a new supercomputer that could be used to solve the toughest problems arising in mathematics and physics. Such problems are called the Grand Challenge Problems of supercomputing. Back in 1989 the year I completed my invention and began appearing in major U.S. newspapers, Seymour Cray —who was the then leading mind in the world of vector processing supercomputers—could not understand the mathematics of where each of my 64 binary thousand initial-boundary value problems should be and when it should be at each of my as many processors. Back in 1989, the supercomputer scientists that attempted to parallel process across an ensemble of processors were processing by the seat of their pants. That is, those supercomputer scientists of the 1970s and ‘80s did not understand the complicated mathematics and did not have the command of scientific materials and the subject matter knowledge that was needed to solve the Grand Challenge Problem. The Grand Challenge Problem was at the crossroad where calculus, algebra, physics, computing, and supercomputing met each other. Because they did not understand the Grand Challenge Problem, the 25,000 vector processing supercomputer scientists of the 1980s and earlier began to hate the massively parallel processing supercomputer that I used to solve the Grand Challenge Problem. Back in the 1980s, I was the only internet scientist that knew the 64 binary thousand, or the 65,536, “address tags” that directed where on my internet each of my 65,536 initial-boundary value problems of calculus and physics was delivered. To the 25,000 vector processing supercomputer scientists of the 1980s, reading my 65,536 email message-passing codes was as incomprehensible as reading a Chinese newspaper. My invention of how to provide the address tags was the necessary pre-condition to the invention of the massively parallel processing supercomputer that is also a new internet de facto. That invention has rich and fertile consequences and contributed to the more complete understanding of how and why the technology called parallel processing—or solving millions upon millions of problems— across as many processors and solving them at the same time makes the computer faster and makes the supercomputer super. 19.1.2 Birth of a New Computer My quest for the precursor to the modern supercomputer that is fastest by parallel processing across a new internet that is a new global network of processors began as a vague idea. That quest began as the seed of an Iroko tree and blossomed, sixteen years later, into the world’s fastest supercomputer that is the Iroko tree of the unknown forest named the massively parallel processing supercomputer. The Iroko tree is the tallest tree in Igbo Land of southeastern Nigeria. The Iroko tree grows along the west coast of Africa. The Iroko tree can live for up to 500 years. My quest for the parallel processing supercomputer began in the early morning of Thursday June 20, 1974 in Corvallis, Oregon, United States. My quest for the modern supercomputer ended at 8:15 in the morning of Tuesday the Fourth of July 1989 in Los Alamos, New Mexico, United States. That quest for the parallel processing supercomputer led to my deeper and surer understanding of the Internet as a planetary supercomputer-hopeful. I had ideas about parallel processing since 1974. But, until the Fourth of July 1989, I did not experimentally prove that parallel processing makes the impossible-to-compute possible-to-compute. My experimental discovery of parallel processing that occurred across my new internet that is a new global network of 65,536 processors occurred on the Fourth of July 1989. Philip Emeagwali Lecture 180913-2

Ask this episode

“What did the council decide on item 42?” — answered with the speaker and the moment.

Transcript

Related