Philip Emeagwali

For the fifteen years, onward of June 20, 1974, I conducted my supercomputer research alone. I did so alone because I was ridiculed, mocked, and rejected by all-white research teams that were exclusively programming only sequential and vector processing supercomputers. As a black African-born supercomputer scientist in the United States, I felt like I was in exile wherever I am. I’m in exile in the United States. I was in exile in Africa. I was in exile in the then uncharted territory of the massively parallel processing supercomputer. A multidisciplinary supercomputer research team could comprise of one thousand scientists and engineers. Each member of that supercomputer research team was at the frontier of knowledge of physics. Or at the frontier of knowledge of mathematics. Or at the frontier of knowledge of computer science. To discover parallel processing required both theory and experiments and required a polymath, rather than a mathematician. To invent the massively parallel processing supercomputer required a polymath that was simultaneously at home at the frontiers of physics, mathematics, and computer science. It took me sixteen years of advanced training, onward of March 25, 1974, in Oregon (United States) as well as weekly attendances at 500 research seminars of the 1980s in the District of Columbia and Maryland (United States), to become that triple threat and that polymath that is at home at the frontiers of knowledge in physics, mathematics, and computer science. Most importantly, I was the only research scientist that gave massively parallel processing research lectures to audiences of research computational physicists at the United States national laboratories. I gave research lectures to research mathematicians at the international congress of mathematicians. I gave research lectures to research computer scientists of the two premier computer societies in the world, namely, The Computer Society of the IEEE and the Association for Computing Machinery. In the late 1970s and early ‘80s, I was rejected because white research scientists dismissed me before they heard me give my research lectures on how I invented the massively parallel processing supercomputer. The audio and video recordings of my lectures on the new supercomputer that I invented are posted at emeagwali dot com. To work cohesively as a supercomputer research team demands that each team member follow the team leader. The supercomputer research teams of the 1970s and ‘80s were coerced to group think and were technologically brainwashed to group think only in the direction of conventional vector processing supercomputing. The leading proponents of vector processing supercomputers were the leading opponents of parallel processing supercomputers. 18.1.4 Sometimes, The Impossible is Possible In 1989, there were 25,000 users of vector processing supercomputers. I was the only fulltime programmer of the handful of massively parallel processing supercomputers of the 1980s. Gene Amdahl and Seymour Cray, the two leading opponents of the parallel processing supercomputer, argued that it will forever remain impossible to parallel process through as many as eight processors or computer cores. In the 1940s through ‘60s, the group thinkers in the field of supercomputing focused only on the sequential processing supercomputer technology. In the 1970s and ‘80s, the group thinkers in supercomputing focused only on the vector processing supercomputer technology. In those two decades, I was forced to work as a lone wolf supercomputer scientist that was not a member of a 400-person research team.