Philip Emeagwali | A Child Soldier of Biafra | Famous Black Inventors of the 20th Century

Philip Emeagwali | A Child Soldier of Biafra | Famous Black Inventors of the 20th Century

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] [Civil War and Corruption in Nigeria] [Early Childhood of Philip Emeagwali] In 1965, I was in the sixth grade
in Saint John’s Primary School, Agbor, Nigeria.
In January 1966, I enrolled in Saint George’s Grammar School, Obinomba,
Nigeria. Fifteen months later,
I fled from Obinomba (Nigeria) to Onitsha (Biafra).
My Igbo-speaking family fled from Nigeria to Biafra
and we fled because thousands of Igbos
from southeastern Nigeria were been killed
in Northern Nigeria. That organized killings of Igbos
occurred from May 29, 1966 through September 29, 1966.
That civil uprising preceded the war between Nigeria and Biafra.
That war began on July 6, 1967 and ended on January 15, 1970.
One in fifteen Biafrans died during that 30-month long war.
In the list of the worst genocidal crimes of the 20th century
that was committed against humanity, the death of one in fifteen Biafrans
was ranked fifth. In the evening of March 21, 1968,
the day my hometown of Onitsha was captured by Nigerian soldiers,
we fled on foot and fled from 14 Mba Road, Onitsha (Biafra)
to Merchants of Light School, Oba (Biafra). Tens of thousands of refugees
that fled from Onitsha were camped
at Merchants of Light School, Oba (Biafra). At about six o’clock
of the following morning of March 22, 1968,
we were alerted by fleeing refugees that advancing Nigerian soldiers
had captured Onitsha and might capture our refugee camp
at Oba and do so within a few hours.
Scared, we continued our flight to Nnewi and Nnobi
and stopped our flight when we reached a refugee camp
that was a former school class room that was across the street
from the Catholic Church in Awka-Etiti (Biafra).
About five days after the war was over, or about January 20, 1970,
we returned as refugees and squatted for five months
in an abandoned house that was along Port Harcourt Road
in the Fegge quarters of Onitsha. In mid-1970, I began to teach myself physics,
algebra, geometry, and calculus. [Struggles Against Corruption in Nigeria] About two weeks
after I received a scholarship letter from Oregon, United States,
that was dated September 10, 1973, I was in Lagos (Nigeria)
to apply for an international travel passport.
Back in 1973, the Nigerian passport or its application forms
cannot be received by mail. At that time, the Nigerian passport office
in Kakawa Street, Lagos (Nigeria) had a reputation
as a cesspool of corruption. All persons applying for
the Nigerian passport spent months coming to the passport office
and did so to monitor the progress of their applications.
Nigerian travel passports were deliberately withheld
by the Chief Passport Officer in Lagos.
Back in 1973, my travel passport was withheld until shortly after Christmas
Day. My passport was withheld
until I paid a bribe of five pounds to one of the passport touts.
I had expected to be in the United States as early as June 1973, at age 18.
I had applied for admission into American schools,
and I applied shortly after, I had passed the entrance examination
to the University of London that I took as an external candidate
back in January 1973 in Onitsha, East-Central State, Nigeria.
My Nigerian travel passport was issued in late December 1973
and after a six month delay. I arrived in the United States
on Sunday March 24, 1974, and after a nine-month delay
and after paying a bribe of five pounds to a passport tout
who claimed that the Chief Passport Officer
gets a large commission from that bribe. That five pounds was a month’s wage.
My Nigerian travel passport was also withheld
until I paid a presumably round trip airfare
from Lagos (Nigeria) to Portland (Oregon, United States).
That two-way airfare was in addition to my one-way airfare
to Portland, Oregon, United States. That two-way airfare
was called “repatriation fee” but it was an extortion fee.
I paid for a round-trip ticket but I was never given any ticket.
I paid 150 pounds, or 30 months salary, as the advance “repatriation fee.”
I paid the Chief Passport Officer in Lagos, Nigeria,
two-and-half years salary, for the privilege of leaving Nigeria
to study in the United States. As a result of that exorbitant extortion
from the Chief Passport Officer, I arrived in the United States
with only 134 dollars, or much less than the bribe
that I paid the corrupt Chief Passport Officer
of Nigeria. I believe that my repatriation fee
went into the personal bank account of the Chief Passport Officer
in Lagos (Nigeria). [Paradigm Shift in Supercomputing] [The Supercomputer in Oregon in 1974] My first night outside Nigeria
was spent in Room 36 of Butler Hall, Monmouth, Oregon, United States.
I checked into Butler Hall at about six in the evening
of Sunday March 24, 1974. Three months later, on June 20, 1974,
I began programming the CDC 3300. That was the first supercomputer
to be rated at one million instructions per second.
That supercomputer was marketed seven years earlier
as the world’s fastest computer. By far, the most important contribution
to the field of supercomputing is to attain a speed
that was once-impossible and then to harness that new speed
to solve the grand challenge problems arising in science and engineering. [A Breakthrough in Supercomputing] Such a breakthrough
in computational mathematics, or the supercomputer solution
of a grand challenge problem, is particularly worthy
of being a benchmark in the history of the computer.
That breakthrough is noteworthy if it changed the way
we looked at the computer and the internet.
With the supercomputer that communicates across processors
and do so synchronously and computes within processors
and do so simultaneously, we now have answers
to previously unanswerable grand challenge questions.
But back in 1974, my unanswerable question
was how to solve a large system of equations of algebra
and how to solve them across a new global network
of 64 binary thousand processors that defined and outlined a new internet.
On June 20, 1974, the day I began programming supercomputers,
the number of computer scientists in the world were few.
That should not come as a surprise. After all, the first computer science academic
programs started only ten years earlier.
For that reason, I was one of only 24 programmers
from around the state of Oregon that were remotely logged into
the supercomputer that was at 1800 SW Campus Way,
Corvallis, Oregon, United States. [New Paradigm in Supercomputing] Three months before I started programming
supercomputers, I had arrived from Onitsha, Nigeria.
It seemed like I was catapulted from a sling shot
from Onitsha to Oregon. At that time, my family in Nigeria
were still struggling to pronounce the word “Oregon.”
The sling shot that catapulted me to Oregon
was a scholarship letter that was dated September 10, 1973.
When I left Nigeria, they was no computer in Nigeria,
or in Sub-Saharan Africa outside of South Africa.
Looking back to 1974, I derived recognition
from being at the frontier of supercomputing
and being there when only twenty-four people
were logged into the primary computer in the entire state of Oregon.
On the sixteenth anniversary of my entry
into the frontier of supercomputing, trade publications
and newspaper articles, such as the June 20, 1990 issue
of The Wall Street Journal, wrote that
I—Philip Emeagwali—had discovered a different way of looking at supercomputers.
I discovered a new paradigm for supercomputing
that uses sixty-five thousand five hundred and thirty-six [65,536]
central processing units to record the once-impossible
3.1 billion calculations per second.
My discovery was a paradigm shift because Seymour Cray—the then leading light in the
world of supercomputers— said that
it will forever remain impossible to use sixty-five thousand
five hundred and thirty-six [65,536] “chickens,” that was his metaphor
for the as many slowest central processing units,
and use them to defeat one strong ox,
that was his metaphor for the fastest
vector supercomputers. [THE PARALLEL SUPERCOMPUTER MAKES THE IMPOSSIBLE
POSSIBLE] I was in the news in 1989
because I discovered that the impossible-to-solve
within a sequential supercomputer is possible-to-solve across
a parallel processing machinery that is not a computer per se.
That new machinery is a virtual supercomputer
and is a new internet de facto. That new internet
is a new global network of sixty-five thousand
five hundred and thirty-six [65,536] central processing units.
At a visceral level, I felt like a nineteen-year-old
that sojourned from the heart of my ancestral Igbo Land
and across the Atlantic Ocean, beyond North America,
and beyond the North Pole and sojourned to reach
the 21st century’s land of the spirits (or ala mmuo), namely,
the unexplored territory of the never-before-seen computer
and the new internet. It was within that unknown world
of the massively parallel supercomputer that I discovered
how to solve the once-impossible grand challenge problems
and thereby extend the boundaries of mathematics, science,
and engineering. I made the impossible-to-solve
possible-to-solve and I accomplished that
when I discovered how to perform the world’s fastest computations
and, far more importantly, discovered how to perform the fastest calculations
and do so with and across the slowest processors in the world. [Contributions to the Development of the Computer] [Philip Emeagwali: What is He Famous For?] In 1989, it made the news headlines
that a lone wolf Nigerian Supercomputer Wizard
in the United States had discovered how to build
the fastest supercomputer and discovered how to always compute fastest.
I am that Nigerian supercomputer scientist
that was in the news back in 1989 and in the news for discovering
practical parallel supercomputing. I was in the news because
I was unconventional and saw something previously unseen,
namely, a new way of supercomputing. In the old way of supercomputing,
a supercomputer that did only one thing at a time
was used to solve the toughest problems that arose in mathematics, science,
and engineering. In my new way of supercomputing,
I used the slowest processors that each merely executed
forty-seven thousand three hundred and three [47,303] calculations
per second per processor.
I am that lone wolf supercomputer scientist
that was in the news for discovering
how to perform the fastest calculations and how to do so across a new internet
that is a new global network of sixty-five thousand
five hundred and thirty-six [65,536] inexpensive, tightly-coupled,
commodity-off-the-shelf processors that shared nothing between each other. What is the contribution
of Philip Emeagwali to the development of the computer? I discovered
how to always perform the world’s fastest computations
and perform it with the world’s slowest
processors. I was in the news, in 1989, because
my experimental discovery of practical parallel supercomputing
marked a milestone in the history of the computer.
For me—Philip Emeagwali— my experimental discovery of 1989
of practical parallel supercomputing wasn’t unexpected.
I expected to confirm my earlier theoretical discovery
of how to massively parallel process across a new internet
that will become a virtual supercomputer.
I expected to confirm that I could communicate across
and compute on sixty-five thousand five hundred and thirty-six [65,536]
computational fluid dynamics codes and communicate and compute them
at once. As a theory, my theoretical discovery
of parallel supercomputing was ridiculed
as a huge waste of everybody’s time. Yet, I discovered
how to save everybody time and how to do so
by synchronously communicating and simultaneously computing
in only one day what used to take
sixty-five thousand five hundred and thirty-six
[65,536] days, or 180 years. [Contribution of Philip Emeagwali to Computer
Development] The contribution
of Philip Emeagwali to the development of the computer
is this: I experimentally discovered
how to parallel process across a new internet
that is a new global network of sixty-five thousand
five hundred and thirty-six [65,536] central processing units.
After my discovery, a grand challenge problem
that formerly took sixty-five thousand five hundred and thirty-six
[65,536] days, or 180 years, of time-to-solution
on one central processing unit now takes only one day
of time-to-solution across a new internet.
Metaphorically speaking, that was how I discovered
180 years in one day. [Why a Supercomputer Scientist Hid His Racial
Identity] Back in 1989, the Award Committee
of The Computer Society was not aware that I was black
and African and for that reason gave me credit
for discovering practical parallel supercomputing
and did so without taking race into consideration.
But scientists that knew that I was black and African
were terribly upset that The Computer Society
gave me the top award in the field of supercomputing
and gave it to me without digging deeper to discover
that I was black and African. In that respect,
the IEEE Computer Society did not give
the top supercomputer award to a black supercomputer scientist.
I simply kept the credits for my contributions
and I could keep them because I was the sole inventor
of practical parallel supercomputing and the sole expert
on the new supercomputer that parallel processed across
my ensemble of 64 binary thousand processors.
Parallel processing appeared as science fiction
on February 1, 1922 and as 64 thousand human computers
working together and in parallel and doing so to forecast the weather.
The precondition to forecasting the weather
is that those 64 thousand human computers
must solve the initial-boundary value grand challenge problem of calculus
that is governed by the primitive equations
of meteorology. For thirty-six years after 1922,
interest in parallel processing was lost, in part, because
the automatic programmable computer that provided the motivation
for faster computing did not exist and was not invented until 1946.
Parallel processing started appearing in computer science literature
and appeared regularly onwards of 1958. For the thirty-one years onward of 1958, parallel
processing was mocked at computer science conferences
and the supercomputer technology was ridiculed as a beautiful theory
that lacked an experimental confirmation. [Changing the Way Mathematicians Count] [Changing the Way We Look at Computers] As a research supercomputer scientist,
my goal is to discover how to compute fastest
and do so with the slowest processors, or how to do more with less
and how to create reality from science-fiction.
Parallel processing—the technology that enables the supercomputer
to solve many problems at once—enabled me
to solve 65,536 problems at once. In principle, your computer
can do whatever my supercomputer can do.
However, your computer that is powered by only one
isolated processor takes 30,000 years to solve a grand challenge problem
that my supercomputer that is powered by an ensemble of
over 10 million processors takes only one day to solve.
Practical parallel supercomputing must be investigated on a broad canvas
and imagined in broad imaginative strokes.
Practical parallel supercomputing only benefits humankind
if and only if it is proven to solve the grand challenge problems.
Practical parallel supercomputing is not for the faint of heart
or for those locked within their own intellectual silos.
As a research supercomputer scientist, my goal was not to merely invent
new algebra and new calculus. My research goal was to project
my new mathematics and project that new knowledge
from the blackboard to the motherboard
and across a new internet that is a new global network of
64 binary thousand processors and, most importantly, to project
that new supercomputer into the real world
where it helps my country of birth, Nigeria, discover and recover
otherwise elusive crude oil and natural gas,
or where it impacts the market trader in my ancestral hometown
of Onitsha. The fastest supercomputer
attracts the toughest mathematical problems
in physics in the manner a high mountain
attracts the storms. The supercomputer is to mathematics
what the Nile is to Egypt. Each is a lifeline.
The supercomputer is an intellectual extension
of the complex equations scribbled on the mathematician’s blackboard.
My goal was to invent a supercomputer out of the slowest processors.
Inventing that supercomputer demanded that I become an athlete
of the mind. Nine out of ten supercomputer cycles
are consumed solving the partial differential equation
of calculus and physics. For that reason,
practical parallel supercomputing may be defined
as solving millions upon millions of initial-boundary value problems
at once. On the Fourth of July 1989,
I announced my discovery of practical parallel supercomputing.
The response from everybody was that I made a mistake.
The first six copies of my 1,057-page research report
that was dated July 4, 1989 that described how I discovered
practical parallel supercomputing were thrown into the dustbin
of the reviewers. I was mocked
and I was warned that I was computing with science-fiction,
not with a new supercomputer. Everybody that said that
I made a mistake was mistaken. Practical parallel supercomputing
has withstood the test of time and is the vital technology
that powers every supercomputer manufactured today.
That experimental discovery that occurred on the Fourth of July 1989
took the parallel supercomputer from a research and development project
to the widespread commercialization that is called the modern computer.
Parallel processing validated the modern computer.
The amount of new computations that I discovered how to compute
on the 4th of July 1989 was 64 binary thousand times
what could be computed only one day earlier. After 1989, massively parallel processing
became the standard technology that must be used in all supercomputers.
Before 1989, the fastest one thousand supercomputers
in the world derived their supercomputing speeds
from only one vector processing unit. After 1989,
the fastest one thousand supercomputers in the world
derived their supercomputing speeds from up to 10.65 million
central processing units that counter-intuitively computed
10.65 million things at once, instead of intuitively computing
only one thing at a time. My 1989 paradigm shift
from computing only one thing at a time to computing 65,536 things at once
opened the door to computing 10.65 million things at once.
A future world without the parallel supercomputer
could be a world without the computer of the future.
If parallel supercomputing is subtracted from human knowledge,
nearly every computer, all supercomputers,
and the internet itself will shut down!! Parallel supercomputing is not
a new knowledge that was created. Parallel supercomputing exists theoretically
and a priori and existed as a technique
that was uncovered for computing faster. I discovered practical parallel supercomputing
when I parallel processed across my new internet
that was a new global network of 65,536 tightly-coupled,
commodity-off-the-shelf processors that shared nothing between each other
and that were equal distances apart from each other.
I turned science-fiction to reality by discovering
how to parallel compute and how to do so sight unseen.
I was in the news back in 1989 because I was the first person
to solve a grand challenge problem and solve it
by massively parallel computing it. I achieved that
supercomputer breakthrough and did so at a time
all my 64 binary thousand processors
were expected to forever remain silent. Parallel supercomputing is an invention because
computers and supercomputers are now parallel processing. Thank you. I’m Philip Emeagwali. [Wild applause and cheering for 17 seconds] Insightful and brilliant lecture

1 Comment

  1. I'm Philip Emeagwali. The contribution of Philip Emeagwali to the development of the computer is this: I experimentally discovered
    how to parallel process across a new internet that is a new global network of sixty-five thousand five hundred and thirty-six [65,536] central processing units. After my discovery, a grand challenge problem that formerly took sixty-five thousand five hundred and thirty-six [65,536] days, or 180 years, of time-to-solution on one central processing unit now takes only one day of time-to-solution across a new internet. Metaphorically speaking, that was how I discovered 180 years in one day.

    Why a Supercomputer Scientist Hid His Racial Identity

    Back in 1989, the Award Committee

    of The Computer Society

    was not aware that I was black

    and African

    and for that reason gave me credit

    for discovering

    practical parallel supercomputing

    and did so

    without taking race into consideration.

    But scientists that knew that

    I was black and African

    were terribly upset

    that The Computer Society

    gave me the top award

    in the field of supercomputing

    and gave it to me without digging deeper

    to discover

    that I was black and African.

    In that respect,

    the IEEE Computer Society

    did not give

    the top supercomputer award

    to a black supercomputer scientist.

    I simply kept the credits

    for my contributions

    and I could keep them because

    I was the sole inventor

    of practical parallel supercomputing

    and the sole expert

    on the new supercomputer

    that parallel processed across

    my ensemble of 64

    binary thousand processors.

    Parallel processing

    appeared as science fiction

    on February 1, 1922

    and as 64 thousand human computers

    working together and in parallel

    and doing so to forecast the weather.

    The precondition

    to forecasting the weather

    is that those 64 thousand

    human computers

    must solve the initial-boundary value

    grand challenge problem of calculus

    that is governed

    by the primitive equations

    of meteorology.

    For thirty-six years after 1922,

    interest in parallel processing

    was lost, in part, because

    the automatic programmable computer that provided the motivation

    for faster computing did not exist

    and was not invented until 1946.

    Parallel processing started appearing

    in computer science literature

    and appeared regularly onwards of 1958.

    For the thirty-one years onward of 1958, parallel processing was mocked

    at computer science conferences

    and the supercomputer technology

    was ridiculed as a beautiful theory

    that lacked

    an experimental confirmation.

    Changing the Way Mathematicians Count

    Changing the Way We Look at Computers

    As a research supercomputer scientist,

    my goal is to discover

    how to compute fastest

    and do so with the slowest processors,

    or how to do more with less

    and how to create reality

    from science-fiction.

    Parallel processing—the technology

    that enables the supercomputer

    to solve many problems

    at once—enabled me

    to solve 65,536 problems at once.

    In principle, your computer

    can do whatever my supercomputer

    can do.

    However, your computer

    that is powered by only one

    isolated processor takes 30,000 years

    to solve a grand challenge problem

    that my supercomputer

    that is powered by an ensemble of

    over 10 million processors

    takes only one day to solve.

    Practical parallel supercomputing

    must be investigated on a broad canvas

    and imagined in

    broad imaginative strokes.

    Practical parallel supercomputing

    only benefits humankind

    if and only if it is proven

    to solve the grand challenge problems.

    Practical parallel supercomputing

    is not for the faint of heart

    or for those locked

    within their own intellectual silos.

    As a research supercomputer scientist,

    my goal was not to merely invent

    new algebra and new calculus.

    My research goal was to project

    my new mathematics

    and project that new knowledge

    from the blackboard

    to the motherboard

    and across a new internet

    that is a new global network of

    64 binary thousand processors

    and, most importantly, to project

    that new supercomputer

    into the real world

    where it helps my country of birth, Nigeria, discover and recover

    otherwise elusive crude oil

    and natural gas,

    or where it impacts the market trader

    in my ancestral hometown

    of Onitsha.

    The fastest supercomputer

    attracts the

    toughest mathematical problems

    in physics

    in the manner a high mountain

    attracts the storms.

    The supercomputer is to mathematics

    what the Nile is to Egypt.

    Each is a lifeline.

    The supercomputer

    is an intellectual extension

    of the complex equations scribbled

    on the mathematician’s blackboard.

    My goal was to invent a supercomputer

    out of the slowest processors.

    Inventing that supercomputer

    demanded that I become an athlete

    of the mind.

    Nine out of ten supercomputer cycles

    are consumed

    solving the partial differential equation

    of calculus and physics.

    For that reason,

    practical parallel supercomputing

    may be defined

    as solving millions upon millions

    of initial-boundary value problems

    at once.

    On the Fourth of July 1989,

    I announced my discovery

    of practical parallel supercomputing.

    The response from everybody

    was that I made a mistake.

    The first six copies

    of my 1,057-page research report

    that was dated July 4, 1989

    that described how I discovered

    practical parallel supercomputing

    were thrown into the dustbin

    of the reviewers.

    I was mocked

    and I was warned that

    I was computing with science-fiction,

    not with a new supercomputer.

    Everybody that said that

    I made a mistake was mistaken.

    Practical parallel supercomputing

    has withstood the test of time

    and is the vital technology

    that powers every supercomputer manufactured today.

    That experimental discovery

    that occurred on the Fourth of July 1989

    took the parallel supercomputer

    from a research and development project

    to the widespread commercialization

    that is called the modern computer.

    Parallel processing

    validated the modern computer.

    The amount of new computations

    that I discovered how to compute

    on the 4th of July 1989

    was 64 binary thousand times

    what could be computed

    only one day earlier.

    After 1989, massively parallel processing

    became the standard technology

    that must be used in all supercomputers.

    Before 1989,

    the fastest one thousand supercomputers

    in the world

    derived their supercomputing speeds

    from only one vector processing unit.

    After 1989,

    the fastest one thousand supercomputers

    in the world

    derived their supercomputing speeds

    from up to 10.65 million

    central processing units

    that counter-intuitively computed

    10.65 million things

    at once, instead of intuitively computing

    only one thing at a time.

    My 1989 paradigm shift

    from computing only one thing at a time

    to computing 65,536 things at once

    opened the door to computing

    10.65 million things at once.

    A future world

    without the parallel supercomputer

    could be a world without

    the computer of the future.

    If parallel supercomputing

    is subtracted from human knowledge,

    nearly every computer,

    all supercomputers,

    and the internet itself will shut down!!

    Parallel supercomputing is not

    a new knowledge that was created.

    Parallel supercomputing exists theoretically and a priori

    and existed as a technique

    that was uncovered for computing faster.

    I discovered practical parallel supercomputing

    when I parallel processed across

    my new internet

    that was a new global network

    of 65,536 tightly-coupled,

    commodity-off-the-shelf processors

    that shared nothing between each other

    and that were equal distances apart

    from each other.

    I turned science-fiction to reality

    by discovering

    how to parallel compute

    and how to do so sight unseen.

    I was in the news back in 1989

    because I was the first person

    to solve a grand challenge problem

    and solve it

    by massively parallel computing it.

    I achieved that

    supercomputer breakthrough

    and did so at a time

    all my 64 binary thousand

    processors

    were expected to forever remain silent.

    Parallel supercomputing is an invention because computers and supercomputers

    are now parallel processing.

    Thank you.

    I’m Philip Emeagwali.

Leave a Reply

Your email address will not be published. Required fields are marked *