« October 1986 | Main | March 2004 »
June 23, 1990
My Eureka Moments
Looking Back
It was an incredible feeling performing computations no one else had ever achieved. I told my wife that my name would be in the newspapers and that I will be famous for doing something nobody has ever done: use 65,000 weak processors to outperform a supercomputer.
Because of internal timer error, I knew that I had to continue until I resolve the timing problem before I could declare my project completed. I almost gave up after I had used 65,536 processors to perform 3.1 billion calculations per second, the calculation that is my claim to fame.
I began my project 30 years, ago and completed it 15 years ago. Traumatized by the racism that I encountered in science, I declared a self-imposed silence on that project. My claim to fame was what I described in a 40-page report published in 1989. However, it took me 15 years to complete the full 1057-page report. The final report was 26 times longer and contains 41 discoveries and inventions.
The misconception is that I had an “Eureka moment,” following a straight line, and that what I had was an overnight success. Far from it. My success was derived from 15 years of continuous study and research, during which many predicted that I would fail.
I devoted the first ten years to conceptual programming of a hypothetical hyperball computer. It had a hypothetical 64,000 computers that were inspired by a 1922 science fiction article.
Conceptual programming is conducted for a computer that has never being built. Ada Lovelace, a 19th century mathematician, was credited as the first programmer even though she only did conceptual programming. The Pentagon of the United States Department of Defense recognized her contributions by naming a programming language Ada.
I did conceptual programming of the hyperball who existed on paper, and practical programming of the hypercube which has
been built.
I then devoted another five years to the practical programming of a hypercube computer with 64 binary thousand processors.
I spent 15 years writing a thousand-page report on how to harness the power of a supercomputer with 64 binary thousand processors. Needless to say, I did not have a user’s manual. No Fortran compiler, no communication primitives, no C compiler, no LISP compiler. The only choice was conceptual programming and then assembly level coding.
For two more years, I repeated the same supercomputer calculations each day. I did not have an accurate internal clock (timer) to measure the speed of my calculation.
LESSONS LEARNED
Life is like a 26.2 mile marathon race. Losers are those that quit at the 26-mile mark. Winners are those that cross the 26.2 mile line. The last 0.2 mile is extremely important and must be completed.
What does it take to get to the top in any field? We have to strive for continuous improvements in our lives. We have to be positively dissatisfied with our achievements.
I have seen the enemy and He is Us.
The story of the risks that I took and the rewards and success that followed resonated deeply within African youths. The message is "If Emeagwali can become a top scientist, so can I."
It does not require above average intelligence to become a scientific genius. Geniuses, after all, are those who learned to create their own positive reinforcements when their experiments were yielding negative results. Perseverance is the key.
Posted by emeagwali at 12:29 PM | Comments (1)