BitDepth 632 - June 17
16/06/08 23:03 Filed in: BitDepth - June 2008
Phillip Emeagwali delivered the inaugaral lecture at the Kwame Ture Lecture Series this year.
Colour among the circuits
Phillip Emeagwali, photographed at the Hyatt hotel. Photo by Mark Lyndersay
Phillip Emeagwali is a smart scientist and something of a rarity in technology circles, a scientist who has done most of his ground breaking work alone, working his way from theory to practice in long 16 hour days of programming.
That he is black should be something that's beside the point of all that work, but it isn't.
Emeagwali was born in Nigeria in 1954, was forced to drop out of school at age 12 and was conscripted into the Biafran army to fight in the war torn nation at 14. He completed his high school equivalency through self-study and came to the United States in 1974 on a scholarship.
Photographs of the young student at the time show a gawky young man with a huge afro and horn-rimmed black glasses in a dashiki, a nerd of colour.
In 1990, he was in San Francisco, the sole black face save for his wife, Dale to receive the 1989 Gordon Bell Prize for his work programming the Los Alamos Supercomputer, winning in the "Price-Performance" category for his work on oil reservoir modelling.
"I'd read a science fiction book years before about a weather prediction system that linked 64,000 minds together, and that set me to thinking about a solution," Emeagwali said at our interview the evening before he spoke at the Kwame Ture Lecture series.
Many processors, one task
"Nobody believed that it was possible to link all those processors together to work together, but I imagined it as a 16 sided hypercube in which each processor would talk to its 16 neighbours in executing the instructions."
Emeagwali set those 65,536 processors humming away on what he described as a "planet-sized problem" the flow of oil beneath the surface of the earth.
His work became part of the foundation of a technology that we now take for granted, parallel computing, the science of setting multiple processors to work on a single set of computing instructions.
In a world in which computers with two processors are almost a norm and quad and octa core systems cost just a few hundred dollars more, it's hard to imagine the world of 1990, when supercomputers were a resource you had to apply for access to and the physical boxes spanned four tennis courts.
It's compelling to note that when Phillip Emeagwali and his wife arrived in San Francisco, he chose to keep his identity quiet until the very moment that his name was announced and he rose to go to receive his award.
"I had some fears and doubts," Emeagwali said. "Would they find some reason to withdraw the award? I don't remember much about what happened, but my wife told me afterward that the room was full of shocked faces, and the presenter almost dropped the award."
Asked about his experience being black on the frontier of technology, there was a surprising response to a question he didn't seem practised in answering.
"You have to deny your black identity," he responded. "If you don't, you end up just being the only black scientist."
Difficult background
There are some troubling aspects to Phillip Emeagwali's career as even a cursory Web search will uncover. There is some doubt regarding the status of his doctorate degree and in this interview, he persisted in describing the Gordon Brown award, a prestigious achievement in its own right, as equivalent to the Nobel prize.
Emeagwali's story cries out for tidying up because the important, unassailably vital material isn't front and centre, his Web site is dated and the repetition of silly claims like being described as the "Bill Gates of Africa" are simply counter productive.
Asked directly about his doctoral degree, Emeagwali responded tangentially, "I've been a student all my life and will remain one. I created new knowledge in the fields of physics, applying Newton's Second Law to oil flows, mathematics, writing the equations to support it and computing, programming 65,536 computers to work on the task. You don't have a discovery until you apply theory to solve a problem."
Phillip Emeagwali, photographed at the Hyatt hotel. Photo by Mark Lyndersay
Phillip Emeagwali is a smart scientist and something of a rarity in technology circles, a scientist who has done most of his ground breaking work alone, working his way from theory to practice in long 16 hour days of programming.
That he is black should be something that's beside the point of all that work, but it isn't.
Emeagwali was born in Nigeria in 1954, was forced to drop out of school at age 12 and was conscripted into the Biafran army to fight in the war torn nation at 14. He completed his high school equivalency through self-study and came to the United States in 1974 on a scholarship.
Photographs of the young student at the time show a gawky young man with a huge afro and horn-rimmed black glasses in a dashiki, a nerd of colour.
In 1990, he was in San Francisco, the sole black face save for his wife, Dale to receive the 1989 Gordon Bell Prize for his work programming the Los Alamos Supercomputer, winning in the "Price-Performance" category for his work on oil reservoir modelling.
"I'd read a science fiction book years before about a weather prediction system that linked 64,000 minds together, and that set me to thinking about a solution," Emeagwali said at our interview the evening before he spoke at the Kwame Ture Lecture series.
Many processors, one task
"Nobody believed that it was possible to link all those processors together to work together, but I imagined it as a 16 sided hypercube in which each processor would talk to its 16 neighbours in executing the instructions."
Emeagwali set those 65,536 processors humming away on what he described as a "planet-sized problem" the flow of oil beneath the surface of the earth.
His work became part of the foundation of a technology that we now take for granted, parallel computing, the science of setting multiple processors to work on a single set of computing instructions.
In a world in which computers with two processors are almost a norm and quad and octa core systems cost just a few hundred dollars more, it's hard to imagine the world of 1990, when supercomputers were a resource you had to apply for access to and the physical boxes spanned four tennis courts.
It's compelling to note that when Phillip Emeagwali and his wife arrived in San Francisco, he chose to keep his identity quiet until the very moment that his name was announced and he rose to go to receive his award.
"I had some fears and doubts," Emeagwali said. "Would they find some reason to withdraw the award? I don't remember much about what happened, but my wife told me afterward that the room was full of shocked faces, and the presenter almost dropped the award."
Asked about his experience being black on the frontier of technology, there was a surprising response to a question he didn't seem practised in answering.
"You have to deny your black identity," he responded. "If you don't, you end up just being the only black scientist."
Difficult background
There are some troubling aspects to Phillip Emeagwali's career as even a cursory Web search will uncover. There is some doubt regarding the status of his doctorate degree and in this interview, he persisted in describing the Gordon Brown award, a prestigious achievement in its own right, as equivalent to the Nobel prize.
Emeagwali's story cries out for tidying up because the important, unassailably vital material isn't front and centre, his Web site is dated and the repetition of silly claims like being described as the "Bill Gates of Africa" are simply counter productive.
Asked directly about his doctoral degree, Emeagwali responded tangentially, "I've been a student all my life and will remain one. I created new knowledge in the fields of physics, applying Newton's Second Law to oil flows, mathematics, writing the equations to support it and computing, programming 65,536 computers to work on the task. You don't have a discovery until you apply theory to solve a problem."
blog comments powered by Disqus
