TIME magazine called him
“the unsung hero behind the Internet.” CNN called him “A Father of the Internet.”
President Bill Clinton called him “one of the great minds of the Information
Age.” He has been voted history’s greatest scientist
of African descent. He is Philip Emeagwali.
He is coming to Trinidad and Tobago to launch the 2008 Kwame Ture lecture series
on Sunday June 8 at the JFK [John F. Kennedy] auditorium
UWI [The University of the West Indies] Saint Augustine 5 p.m.
The Emancipation Support Committee invites you to come and hear this inspirational
mind address the theme:
“Crossing New Frontiers to Conquer Today’s Challenges.”
This lecture is one you cannot afford to miss. Admission is free.
So be there on Sunday June 8 5 p.m.
at the JFK auditorium UWI St. Augustine. [Wild applause and cheering for 22 seconds] [Turning Mathematical Fiction to Supercomputer
Reality] [Supercomputer Science-Fiction] Back in the 1970s and ‘80s,
to parallel process was to cross the border
between the known vector supercomputer and the unknown ensemble of
64 binary thousand processors that was a supercomputer-hopeful.
The upper limit of my quest for the fastest computation
was my parallel processed supercomputing in the 64th mathematical dimension
in which two-raised-to-power-64 processors that were identical to each other
had a one-to-one correspondence with the vertices of the cube
in the 64th dimensional hyperspace. That upper limit in parallel processing
will remain in the realm of science fiction. [What Makes a Supercomputer Super?] The word “computer”
had different meanings to each generation.
To most people, the laptop is the computer.
Back in the 1980s, the desktop was the computer.
However, one thing that has not changed is the definition of the computer
as a machinery that at its core executes fast calculations.
The supercomputer is any of the one thousand
fastest computers in the world. I am often asked:
“What makes a supercomputer super?” For me, the new supercomputer
that I invented was a new global network of processors
that had no central control processor. The new supercomputer
that I invented is a new internet because
it executes its calculations across a new global network of processors.
The fastest parallel processed computations and communications
could only be experimentally discovered on the cusp
between the dream planetary-sized supercomputer
and tomorrow’s science fiction internet. I was the subject of school reports because
I discovered how to evenly divide
real world grand challenge problems and discovered
how to map those real world problems and how to distribute them
with a one-problem to one-processor correspondence and how to simultaneously solve
those problems across millions upon millions
of commodity processors that were identical to each other
and that shared nothing between each other.
I emailed each smaller problem as a digital code of zeroes (0s)
and ones (1s). I divided each grand challenge problem according
to a set of rules. I gave each emailed code
a header that described which processor the code is from.
That header also described which processor should receive
the code and where the code belongs
in the grand challenge problem that is an ensemble of millions of
smaller computational physics problems that each is an
initial-boundary value problem that is governed by my system of
partial differential equations of calculus.
The new mathematical knowledge that I just described
is the mathematical essence of the Philip Emeagwali Formula
for the world’s fastest computer that then U.S. President Bill Clinton
described in his White House speech of August 26, 2000
that made the news headlines. That new knowledge—called
practical parallel processing that I discovered
on the Fourth of July 1989— was what made the supercomputer super? [Quest for Human Progress] The United States constitution
is amended, occasionally, to bring it up to 21st century reality.
My supercomputer lectures must be similarly amended
to bring them up to date. In particular, I had to enlarge
what it means to smooth out the jagged frontiers
of scientific knowledge and to contribute
in the technological context of 21st century computer science.
The quest for human progress is a journey to the future.
And to the terra incognita where the scientific discovery
is the magical act of showing that, sometimes, that believed
to be impossible is, in fact, possible. To invent—or see something
that was previously unseen— is to create the future.
It’s like doing something no human had done before.
Or like travelling to a planet no human had visited before you. [My Quest for the World’s Fastest Computer] [My Early Years as a Black Supercomputer Geek] I define the supercomputer
as any computer that is listed within
the top one thousand fastest computers in the world.
By my definition, the few computers of the 1940s, ‘50s, and ‘60s
were supercomputers. And the computers that I programmed
in the 1970s and ‘80s were supercomputers.
Retrospectively and as a sub-Saharan African-born scientist
in the United States who came of age in the 1970s and ‘80s, my
scientific career took a path that some thought
it should not have taken. Back in 1989, many people struggled to understand
why a black man was the sole fulltime programmer
of the most massively parallel supercomputers ever built.
The answer, in part, is that I started programming—the CDC 3300—one
of the world’s fastest supercomputers back on June 20, 1974
at 1800 SW Campus Way, Corvallis, Oregon, United States.
I remember that date as eighteen days before President Richard Nixon
was forced out of The White House. The maximum of eighty
computer programmers—at a time and from the entire state of Oregon—indicates
that they were only a few hundred computer programmers
in Oregon in 1974. [What’s It Like Being the Only Black Supercomputer
Scientist] Back in the 1980s
and within the nuclear research laboratories in the United States,
where active supercomputing research is conducted,
I was treated like a “security threat.” I was de facto an illegal alien
who sought refuge at the frontier of
supercomputing knowledge. For the record,
my earliest supercomputer accounts were revoked whenever it was discovered that
I was black and sub-Saharan African! Because my supercomputer accounts were revoked,
my survival strategy was to stay low key
and do so during my first sixteen years as a supercomputer scientist.
As a black and African research supercomputer scientist
in Corvallis, Oregon, my quintessential question was this: “What did my isolating identity
do to me as a research scientist?” In Corvallis, Oregon,
I lived a very isolating identity and I grappled with existential issues.
After sixteen years of unrecognized supercomputer research,
I began to wonder if one day my contributions
will be forgotten? Being the first person
to be referred to as a supercomputer scientist
confused a lot of people and did so, in part, because
I was black and African-born. As a black parallel supercomputer scientist,
I was mocked and made fun of because I worked alone
and tried to turn the science-fiction of parallel processing
into the non-fiction that is today’s supercomputer.
Back in 1974, Kidder Hall —the symbol of mathematics
in Corvallis, Oregon, United States— was a seemingly majestic structure.
So were the physics and the engineering buildings.
Back in 1974 in Corvallis, Oregon, the computer science department
was hinged in a hastily put together trailer.
In 1974, I didn’t see a future in the field of computer science
because it lacked the respectability to be housed in a multi-story
concrete building. In May and June 1975,
I lived at 15 Edgewood Way, Corvallis, Oregon, United States
That was the residence of Ted and Connie Foulke.
Ted was a chemical engineer that retrained as a physician.
And Connie was a high school teacher. From March 1975 through June 1977,
I parked my red two-speed bicycle at the back of Kidder Hall
at 2000 SW Campus Way, Corvallis, that was 190 feet from
the supercomputer that I was programming
that was the world’s fastest computer when it was manufactured
back in December 1965. From October 1975
through January 1976, I lived at 2540 SW Whiteside Drive,
Corvallis, Oregon, United States. That was the residence of
Fred and Anne Merryfield. Fred was a noted civil engineer
who co-founded a global engineering company
called CH2M. I rode my red two-speed bicycle
that I bought for ten dollars to 2000 SW Campus Way, Corvallis, Oregon,
a distance of 2.6 miles, where I used the Teletype
to access and program the first supercomputer
to be rated at one million instructions per second
and that was 190 feet away and across the street. [A Mathematician’s Quest For the World’s
Fastest Supercomputer] The essence of my existence
is abstract mathematics that is impenetrable
to most research scientists. And solving grand challenge
initial-boundary value mathematical problems
and solving them by massively parallel supercomputing
their algebraic approximations was also impenetrable
to the most able research mathematicians.
Back in 1941, the largest system of equations
of algebra that could be solved
involved only 29 unknowns. The ENIAC and UNIVAC supercomputers came along
in 1946 and ’51, respectively. The CDC 3300 supercomputer
that I programmed in Corvallis, Oregon was introduced in December 1965
and in that year it was the fastest computer
in the world, or the number one ranked supercomputer. The CDC 3300 supercomputer
was used to forecast the weather. [Fastest Possible Computers] The theorized parallel supercomputing
that I invented in my head, back in the 1970s,
was different from the practical parallel supercomputing
that I invented, later in the 1980s. Through two decades of trial and error,
I learned that I could only invent the parallel supercomputer
that could be invented, or rather the fastest supercomputer
that the laws of physics permit me to invent.
Prior to the Fourth of July 1989, the parallel supercomputer
was a technology that I knew but cannot explain
or confirm by an experiment. [Who is a Supercomputer Scientist?] I was the first person to be referred to
as a supercomputer scientist. Supercomputing is a broad field.
I am a supercomputer scientist that placed his emphasis on the science.
Some supercomputer scientists are “mathematicians” who prove
which abstract supercomputer can or cannot solve
a grand challenge problem. Some supercomputer scientists
are “engineers” who build supercomputers. Some supercomputer scientists
are “inventors” who measure the speed of never-before-seen supercomputers
and tries to invent the fastest supercomputer
that is powered by new technologies. I was trained as a research mathematician,
research engineer, and research physicist.
I conducted my supercomputing research at the frontiers
and at the cross road where mathematics, physics,
and computing met. [The Grand Challenge of Supercomputing] In the 1980s, the United States Department
of Energy compiled a list of twenty
impossible-to-solve problems that were very important.
Those twenty problems were, thereafter, dubbed the twenty grand challenges
of supercomputing. Those twenty grand challenge problems were
to computing what the seven millennium problems were to
mathematics. The reason the grand challenge problems that
pertained to physics were exceptionally difficult was that each
problem can only be solved by a polymath
who has command and mastery of physics, algebra, calculus,
and computer science. My parallel processed solution
of the grand challenge problem was highlighted
in the June 20, 1990 issue of The Wall Street Journal
and was described as cover stories of top mathematical publications.
My complete solution was described in my very lengthy series of
online lectures. Only a polymath
will have the confidence to tackle the grand challenge problem. [The Most Beautiful Theory in Computer Science] Back in the 1970s and ’80s,
extreme-scale computational mathematicians
didn’t deem the parallel processing of initial-boundary value problems
of mathematical physics as merely difficult.
Mathematicians deemed the parallel processing of real-world grand
challenge problems as impossible. For that reason, it was then said that parallel
processing was a beautiful theory
that lacked an experimental confirmation.
My contribution to the development of the supercomputer is this:
I provided that lockdown experimental confirmation
of parallel supercomputing and I made that discovery
at 8:15 in the morning of the Fourth of July 1989
in Los Alamos, New Mexico, United States.
I remember that date because it was the U.S. Independence Day.
I discovered parallel supercomputing and I discovered
the supercomputer technology across my ensemble of 65,536 processors
that was the precursor to the current world record
of 10.65 million processors. Thank you. I’m Philip Emeagwali. [Wild applause and cheering for 17 seconds] Insightful and brilliant lecture