All I Really Need to Know
I Learned in Kindergarten
An interview with the National
Society of Black Engineers
Interviewer: Ed Woodhouse (February 19, 2002).
National Society of Black Engineers:
As a child in Western Nigeria, your father would indulge you in a drill where you would have to complete 100 math problems in one hour. How has that drill helped you in your work with computers?
Emeagwali:
The essence of that drill was "fast computation." My research is on "fast computations." The essence remains the same while the technique has changed. The old technique was that I used my human brain to solve 100 math problems in one hour. The new technique is that I used 65,000 electronic brains to solve 3.1 billion math problems in a second. It reminds me of the popular saying "All I really need to know I learned in kindergarten" and "The more things change, the more they remain the same."
As an adult mathematician, I am still solving the same type of problem that I solved as a child in Nigeria. I begin with partial differential equations and reduce them to billions of arithmetical operations, which are similar to the ones that I solved as a child in Nigeria.
The poet T.S. Elliott was right when he wrote that: "We must not cease from exploration. The end of all our exploring will be to arrive where we began and to know the place for the first time."
National Society of Black Engineers:
Who or what motivated you to one day make a better life for you and your family?
Emeagwali:
It is very easy to become motivated if you lived in one of the harshest places in the world and become aware that a good education offers an opportunity for a better life. The struggle that I had in my early years in Africa is what motivates me.
I quit school at the age of 12 because my family was forced to flee during an ethnic cleansing in which 50,000 Igbo-speaking Nigerians were killed. After living in a refugee camp for two years, I was conscripted into the Biafran army as a child-soldier in one of Africa's bloodiest conflicts.
I grew up in a household of ten that survived on less than one dollar a day. Our life in Africa was very low tech and simple tasks took huge amounts of time. I go to the forest to fetch firewood to cook, walk to the river to get drinking water, and I walked miles to school, if I was lucky to be enrolled. Children of Africa are used to working hard and that translates to anything they do later in life.
National Society of Black Engineers:
Would you please share with our readers the influence a science fiction article on how to use 64,000 mathematicians to forecast the weather for the whole Earth inspired you?
Emeagwali:
The essence of that article was that many computation-intensive problems could be solved by harnessing the power of 64,000 human computers. That work inspired me to invent an international network for 64,000 far-flung electronic computers that are uniformly distributed around the Earth. I argued that it could be used to forecast the weather for the whole Earth but it was rejected. "For your own good do not publish it," I was advised.
To make it more practical and publishable, I revised it from "64,000 far-flung computers connected as a HyperBall" to 64,000 computers interconnected as a hypercube. The experts believed that it will be impossible to program thousands of processors forecast the weather and they told me: "There's no way it'll work." I remember an article in The New York Times (11/29/89) in which a leading expert said: "We can't find any real progress in harnessing the power of thousands of processors."
A few weeks later, my use of 64 binary-thousand processors contradicted the argument that we could not "harness the power of thousands of processors." The publicity lead to a rediscovery of my original HyperBall scheme.
National Society of Black Engineers:
You submitted a proposal many years ago to the Los Alamos officials remotely because you didn't think the lab officials would accept it if they knew you were black. Can you share with us the details of that situation?
Emeagwali:
Because I had a supercomputer account revoked when it was discovered that I am black, I decided to conduct my research remotely at a dozen different places. Since I mainly communicated by email, and occasionally by telephone, the officials were surprised to discover from the news reports that I was black. If the officials knew that I was black there would have questioned the wisdom of my attempting to solve a problem that defied the brightest white mathematicians.
My final calculations were conducted on an experimental machine belonging to the Los Alamos National Laboratory, the birthplace of the atomic bomb and the headquarters of supercomputing.
At that time, the United States supported the "Comprehensive Nuclear-Test Ban Treaty," which will prohibit "any nuclear weapon test explosion or any other nuclear explosion." Since it will be illegal to test atomic bombs in a desert or an ocean, the nuclear weapons designers at Los Alamos were instructed to figure out how to harness the power of thousands of processors to simulate nuclear explosions.
In my proposal, I explained that there is no distinction between the mathematical equations used in nuclear explosions and weather forecasting, except that one is classified and the other non-classified.
In principle, the weapon designers and I were performing similar calculations: temperature, pressure, and density in a nuclear weapon and petroleum reservoir, respectively. The small difference is that a nuclear explosion takes place in an instant (or what mathematicians describe as small time and space scales) while a petroleum reservoir is operated for about 20 years. Also, the goal is to ensure that the conditions within a nuclear weapon are extreme. It is the extreme temperatures and pressures of a nuclear weapon that obliterates everything within its path.
Because I had completed a thousand-page monograph on how to use 65,536 processors to solve partial differential equations, I reasoned that it will make no difference whether the equations were for nuclear explosions or weather forecasting or oil reservoir simulation. As a physicist, I knew that at the core, a nuclear explosion is about motion and therefore, must be governed by the same laws of motion that govern flows within the Earth's atmosphere and oilfields.
Los Alamos officials understood that if I succeed in using 65,536 processors to simulate an oilfield that their weapons designers could use the same technique to simulate nuclear blasts.
Therefore, my work was a valid proof-of-principle for nuclear weapons designers. The officials purchased two identical supercomputers: one classified and the other unclassified. I used the unclassified supercomputer while the weapons designers used the classified one. In exchange for using their computers, the weapons designers could study, learn and take ideas from my codes. We had a one-way mirror, they could examine my code but I would have been jailed for espionage if I did the same to them.
National Society of Black Engineers:
How much you think things have changed for minorities in the science and technology field for those potential trendsetters such as yourself who are also minorities?
Emeagwali:
Remember that Africans were not always minorities in science. The book "Biographical Encyclopedia of Science and Technology" written by Isaac Asimov, acknowledges that the first known scientists were Africans.
Asimov named Imhotep, a full-blooded Negro African, as the first known scientist, medical doctor, engineer and architect. The second known scientist is an African named Ahmes, who wrote the world's oldest mathematics textbook. The pioneers of science were born, raised, and educated in Africa. Therefore, the ancients did not doubt the ability of Africans to contribute to science. According to Asimov's book, technology is the gift of ancient Africans to our modern world.
To justify the institution of slavery, the myth was created that Africans are intellectual inferior to Europeans. For example, Thomas Jefferson wrote in his book "Notes on Virginia" that it will be impossible for an African to understand high school geometry.
When Benjamin Banneker forwarded his mathematical calculations to him, surprised Jefferson forwarded it to the French Academy of Sciences. The Academy informed Jefferson that Banneker is a mathematical genius and that was how he became famous.
Two hundred years later, I was reliving Banneker's experience. The more things change the more they remain the same.
National Society of Black Engineers:
You describe yourself as a black scientist with a social responsibility to communicate science to the black diaspora. Will those responsibilities you feel have to be extended to all races and cultures due to the global effect computers have had on the world? Please elaborate.
Emeagwali:
The Internet has always been a primary means of communication for me. The scientific community discovered me on the Internet. The public is now rediscovering me on the Internet. Since the Internet is global, it makes sense to reach out to all races and cultures.
The three I's principles that guide the development of my Web site are: Inform, Inspire and Involve. I avoid the dry, information only approach used by scholars.
National Society of Black Engineers:
Do you think the equation of genius with science has stagnated blacks wanting to pursue careers in the science field, fearing they may not be smart enough? Please elaborate on your response.
Emeagwali:
The world's greatest mathematicians and physicists graduated at the bottom of their classes. Isaac Newton was at the bottom of his class at Cambridge University. So was Albert Einstein. Both Newton and Einstein are considered the two greatest mathematical physicists of all times. Any person that believes that she is not "smart enough" should first think of Newton and Einstein.
National Society of Black Engineers:
Mathematics was your favorite subject in school. What do you feel should be the focus of today's school curriculum for elementary to high school students to supplement the demand and necessity of the computer?
Emeagwali:
Until recently, the Swedish called the computer "matematikmaskin" or "mathematics machine." The computer is the child of mathematics and we cannot separate computing and mathematics. The early computer pioneers were trained solely as a mathematician. I was a mathematician until computer scientists discovered me, adopted me, and reintroduced me to mathematicians.
Trying to study computer science without a good knowledge of mathematics is like attempting to play soccer without learning how to run.
National Society of Black Engineers:
What is your definition of computers?
Emeagwali:
The word "computer" has been around for six centuries and meant different things to different generations. The reason is that the computer has been reinvented over and over and its definition has, consequently, changed over the centuries.
In the 14th century, a computer meant a person that calculates time. In the 19th century, it meant a person that computes. In the 20th century, it meant a machine that uses one processor to compute.
In the 21st century, it will be redefined as a machine that uses thousands of processors to compute. This is my definition and the major conceptual leap in the field of computing that I will predict. My definition is different because the computer has been improved and reinvented to use multiple processors.
My definition differs from that of Alan Turing, who introduced the concept of a universal programmable computer that can store and process information as binary digits within a single processor. Instead of using a single processor, I used thousands to simultaneously store, compute and communicate.
It also differs from that of John von Neumann, who proposed that we store a program in one memory. In my work, I store my programs in thousands of distinct memories.
The essential difference between my approach and that of Turing or Neumann is that they used one processor (and memory) while I used 65,000 processors, each with its own distinct memory, to achieve a 65,000-fold increase in computational speed. Because my 65,000 processors must communicate, I had to both compute and communicate.
By the end of this century, I predict, that computers and the Internet will be both obsolete and that we will remove those words from our daily lexicon.
National Society of Black Engineers:
Have things progressed to your satisfaction in recent years in regard to the white male dominated scientific establishment that you've said at times excludes African Americans from succeeding in scientific fields?
Emeagwali:
In the mid 1980s, I interviewed at dozens of leading government research laboratories. They were surprised when I showed up at the interviews. The interviews ended with: "Thank you, we will call you later." The phone call never came and I accumulated a great pile of Dear John letters.
I learned a truism in American science: "if you're black, stay back. If you're brown, stick around. If you're white, you're alright." White scientists had this belief that the lighter skin you are, the more intelligent you will be and the more opportunities should be offered to you. As a result of getting more opportunities, you will find that the list prominent black scientists from Benjamin Banneker to Charles Drew were light skinned.
I realized that the chances of my conducting research in any government lab were slim and none. My first offers came after I became renowned than the white scientists who would have acted as my supervisors.
Exclusion of African-Americans from scientific research is a vestige from the times that it was illegal to teach a black child how to read and write. In the mid 19th century power belonged to those who were literate and it was made illegal for a black child to get an education in America. Secret schools were operated for slaves. Black children had to hide their books from the police. White people understood that it will be difficult to control an educated black person and that such people are likely to revolt.
Similarly, in the 21st century, the power belongs to those that possess technological knowledge and white scientists feel threatened by the presence of African-Americans. White dominance in science is cited as proof of white supremacy and the best way to maintain that dominance is by excluding blacks from paid scientific jobs.
National Society of Black Engineers:
Would you please expand on your theory of one of the many ways the economy can be strengthen is to place a ban on building new prisons and converting existing prisons to degree-granted colleges?
Emeagwali:
California and Florida spend more in prisons than in higher education. Nationwide, we are building more cell blocks and fewer classrooms. It costs $54,000 to build a cellblock. Since the money is borrowed from a bank, the compounding interest will mean that the true cost will be about $100,000 for each cellblock. It costs $25,000 a year to operate each prison bed. Add the police, legal and court costs that preceded the prison sentencing.
These inmates cannot pay taxes or provide child support. Many are nonviolent young men that were sentenced for buying a few ounces of cocaine. They have learned their lessons and want to stay straight and work for a living. We do not need to keep them in prison for several years.
What we have is a vicious cycle: reducing the education budget increases crime and requires us to build more prisons. It makes more sense for us to build fewer prisons and redirect the savings to education. To reduce education budget is nothing but pennywise and pound foolish.
Racism is the fuel that drives the boom in prison cell construction. Sixty percent of drug offenders incarcerated in state prisons are African-Americans. Yet only 11 percent of drug users are African-Americans. One in three young black man is under correctional control and supervision and, in the early 1990s, there were more blacks in prison than in colleges.
National Society of Black Engineers:
What are your feelings or concerns about Apple Computers falsely claiming to have reinvented the computer by incorporating your multiprocessor technology in their four-processor PC benchmark at 3.1 billion calculations per second?
Emeagwali:
In 1988, many thought I had made a mistake when I announced that I had used 65,000 processors to perform the world's fastest computation of 3.1 billion calculations per second. Apple doing the same with fewer processors validates my results and Steve Jobs has no right to claim that he has reinvented the PC. The multiprocessor technology enabled Apple Computers to double their performance by merely doubling the number of processors.
Although, the word "computer" was coined 600 years ago, the computer has been reinvented over and over. The terminology is the same. The technology is different.
Reinventing computers is a process that has been going on for thousands of years. The abacus was reinvented as a Table of Logarithms. The logarithm was reinvented as a slide rule. The slide rule was reinvented as mechanical computer. In the 1940s, the mechanical computer was reinvented as the digital computer.
Finally, we have reinvented the microprocessor to multiprocessor. I demonstrated that a computer that uses thousands of inexpensive processors is more powerful than any other supercomputer. In that sense, my work contributed to the reinvention of the computer. As a result, all supercomputers use thousands of processors.
The supercomputer has been reinvented and the supercomputer of today is the computer of tomorrow. Therefore, we are in the process of reinventing the computer of tomorrow.
Multiprocessing technology has hundreds of applications, besides the PC. Remember that in December 2000 the media reported that Saddam Hussein has cornered the PlayStation market, making it difficult for parents to purchase them as Christmas gifts. Saddam Hussein became the Grinch that stole Christmas 2000 from many children.
Saddam Hussein wanted Iraqi computer scientists
to strip the processors from the PlayStations, reinstall Linux
operating system, reprogram the coprocessors, link them up as
more powerful computers which will guide and deliver long-range
missiles and nuclear bombs. Saddam's new guided missiles could
deliver biological germs with greater accuracy than the earlier
SCUD missiles that struck Jerusalem during the Persian Gulf War.
He believes that a nuclear bomb will enable him to stand up and
so NO to American aggression.
National Society of Black Engineers:
Of all the awards and honors you've received over your illustrious career, which award or honor are you most proud of and why?
Emeagwali:
The 1989 Gordon Bell Prize, considered computation's Nobel Prize.
I am proud of it because it was a seal of approval from the computing community. It came at a time that I was struggling in obscurity and attempting to gain legitimacy. All the top computer scientists believed that I must have made a mistake in my work. "It's a waste of time," proclaimed Gene Amdahl, IBM's top computer designer.
The Gordon Bell Prize was public proclamation that the experts understood my mathematical equations and computer programs and believed that they are major contributions to knowledge. When a work is technically advanced, few scientists could understand it and its acceptance is based upon the reputation and records of those who claim to understand what I accomplished.
There is a saying in Hollywood that "it takes 15 years to become an overnight success." My work was rejected for 15 years. The first five years I was laughed at. The second five years I was forced to work without pay. The third five years I was paid $750 a month.
The struggle is always harder for a black scientist. I walked through a path strewn with thorns while white scientists walked through ones strewed with roses. The destination may be the same but the black scientist walked the farthest and climbed the highest mountains.
National Society of Black Engineers:
For the sake of those (few they may be) who are not familiar with the particulars of your work as a mathematician and inventor, can you briefly talk about your work, and why you are called "A Father of the Internet"?
Emeagwali:
Science grows by accumulation of knowledge. The Internet does not have one birth date or place. Since the Internet is the product of a succession of inventions that took place over several decades, recognition of any individual as "A Father of the Internet" should be qualified by stating that the Internet has many fathers, mothers, uncles and aunts.
My contribution, according to the book "History of the Internet," is that I invented the HyperBall scheme for a supercomputer so large that it covers the whole world.
My HyperBall was a spherical network with 64,000 far-flung processors that will be evenly distributed around the Earth. The Internet is also a spherical network of millions of far-flung computers that are randomly distributed around the Earth.
My work was inspired by a 1922 science fiction article on how to use 64,000 mathematicians to forecast the weather for the whole Earth. However, my proposal to use 64,000 computers to form an international network was dismissed on the grounds that "It will not work." In 1987, I completed a thousand-page monograph that described both the Hypercube and HyperBall topologies, each with 65,536 processing nodes, for performing the world's fastest computation.
When my hypercube implementation won the Gordon Bell Prize, it gave credibility and renewed interest in my HyperBall proposal to use far-flung computers to forecast the weather for the whole Earth.
Because the topology of my HyperBall international network of computers was similar to, but predated that, of the Internet, it was rediscovered and called an "idea that was ahead of its time" and "a germinal seed of the Internet."
National Society of Black Engineers:
What project(s) are you currently involved in?
Emeagwali: I have an ongoing project for a SuperBrain that can be a metaphor for the global network of humans connected to computers and the Internet. In the future, I expect SuperBrain to replace the Internet. I believe that a Super Being will emerge from SuperBrain. Some theologians are fascinated by this research and I recently went to France to discuss how we could extend this work to prove that God exists.
For more information, please contact Dr. Donita Brown at 443-850-0850; fax/voicemail 413-521-3764; email donita@emeagwali.com
back to the web page Philip Emeagwali - Computer Scientist of the African Diaspora
| This website was created by and is maintained
by Dr. Scott Williams, Professor of Mathematics State University of New York at Buffalo |
|
|
|
|
