On Inventing a New Computer Science | Philip Emeagwali | Technology Futurist Speakers


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] [Philip Emeagwali Computer] [Inventing a New Supercomputer] Thank you. Thank you. Thank you very much. I’m Philip Emeagwali. Parallel supercomputing
is the most important invention in the history of the computer. The experimental invention
of the massively parallel supercomputer—that solves many problems at once, instead of solving
only one problem at a time— is one of the computing industry’s
most hopeful narrative. Parallel supercomputing
is an entirely new approach to modern computer science. Without parallel supercomputing,
we will only know what might happen. With parallel supercomputing,
we know what will happen. For two hundred millennia,
we strove to make our world a more knowledgeable place. We discovered new fields of study. The field of study that I discovered,
in the 1970s and ‘80s, is called massively parallel supercomputing. I discovered
the supercomputer-hopeful’s most well-guarded secret, namely, practical
parallel processing across a new internet. But the heart of supercomputing
isn’t the fastest calculations. The heart of supercomputing
is solving the grand challenge problems of computer science. [Changing the Way We Look at the Computer] [On Becoming the First Supercomputer Scientist] I have three [quote unquote]
“first” supercomputers that corresponded to three technologies. The first [quote unquote] “first” supercomputer
of 1946 was the first that was programmable. The second [quote unquote] “first” supercomputer
of December 1965 was the first supercomputer
to be rated at one million instructions per second. I began programming
that second [quote unquote] “first” supercomputer
on June 20, 1974 in Corvallis, Oregon, United States. The third and most important “first” supercomputer
is the precursor to the modern supercomputer of today
that can execute large-scale, excruciatingly detailed
computational fluid dynamics codes, such as climate models,
and execute them across millions of tightly coupled
commodity-off-the-shelf processors. [Serial Processing Versus Parallel Processing
Debate] Back in the 1980s,
parallel processing was controversial and was mocked
by the community of 25,000 vector supercomputer scientists. Looking back, that widespread
rejection of parallel processing gave it’s acceptance in 1989
some street cred. If you invent something
that everybody accepted then you’ve not invented anything new. None of the 25,000 vector supercomputer scientists,
or their leader, Seymour Cray, had a deep understanding
of practical parallel processing. If they understood
how to solve real world problems, as opposed to textbook problems,
and if they understood how to parallel process and how to solve
grand challenge problems and how to solve them across
an actual ensemble of sixty-four binary thousand processors,
then Seymour Cray would not have mocked, ridiculed,
and dismissed parallel processing as a huge waste of everybody’s time. Seymour Cray
is best remembered for his famous quote that ridiculed parallel processing
as the technology that will render his supercomputers obsolete. Seymour Cray,
taunted aspiring parallel supercomputer scientists by asking them:
[quote] “If you were plowing a field,
which would you rather use? Two strong oxen or 1,024 chickens?” [unquote]
This famous quote of Seymour Cray was used to taunt me
when I was attempting to solve the toughest problems
arising in mathematics and physics and attempting to solve them
by parallel processing them across my new internet
that was a new global network of 65,536 processors. Back in 1989,
it made the news headlines that an African supercomputer wizard
in the United States had figured out
that 65,536 chickens, that were his metaphors
for his as many commodity-off-the-shelf processors that were tightly-coupled to each other,
and figured out that those processors can communicate and compute together and do
so as one cohesive whole virtual supercomputer
and do so to become more powerful than one strong ox
that was Seymour Cray’s metaphor for the vector supercomputer. I’m Philip Emeagwali
and I’m the African supercomputer scientist
that was in the news back in 1989. I was in the news because
I discovered that practical parallel supercomputing
will become the vital technology that underpins every supercomputer. [Team Versus Individual Contributions] The comparison of my contributions
during the 1970s and ‘80s, to the collective contributions
of the other 25,000 supercomputer scientists
is an act that lacks historical context and perspective, namely,
my discovery of parallel supercomputing
and comparing that single person’s contribution to the collective contribution
of the one thousand employees of a supercomputer giant
that was paid billions of dollars to deliver supercomputers
that embodied my new knowledge of parallel supercomputing. [Mount Everest of Science] I began programming
one of the world’s fastest supercomputers back on June 20, 1974, at age nineteen,
in Corvallis, Oregon, United States. What is a supercomputer? The fastest supercomputer
occupies the space of a soccer field. The fastest supercomputer
costs the budget of any of the forty poorest nations in the world. The supercomputer is like the lighthouse
for science. Supercomputing
at the fastest recorded speed is like climbing
the Mount Everest of science. [Inventing a New Computer Science] [Inventing the World’s Fastest Computer] How fast is a supercomputer,
compared to a computer? Supercomputing at the speed of
one billion trillion calculations per second
is like if each of the world’s eight billion persons
solved 125 billion math problems per second. If one million
human supercomputers computed at the speed of
one calculation per second, they will take 1.5 billion years
to complete a calculation that takes only one second
on a parallel supercomputer that computed at the speed of
one trillion billion calculations per second. [Chronology of Contributions of Philip Emeagwali] My contributions to science, namely,
practical parallel supercomputing, that occurred on the Fourth of July 1989
and occurred at age thirty-four (34) should be measured
by a different yardstick when comparing them
with my cumulative scientific discoveries at age sixty-four (64)
or later at age ninety-four (94). The reason is that in thirty (30) years,
or sixty (60) years, I will have doubled, or tripled, my age
and developed a larger body of discoveries and inventions. Comparing my contributions
at ages thirty-four (34), sixty-four (64), and ninety-four (94)
is like comparing the gold medal counts at the Olympic Games
of different sized nations, such as Liberia, Nigeria,
and the United States. In scientific research,
the number of papers published doubles every nine years. My quest was for the discovery of practical
parallel processing that is the vital technology
that changed the way we looked at the supercomputer. Contrary to a widely held misconception, an
article is not a contribution to science per se. The fifty million scholarly articles published
in year 2010 alone are nothing but distant
background noises. [Inventing a New Computer Science] The Grand Challenge problem
of mathematics and physics was solved by inventing
a new computer science, called practical parallel processing. It’s impossible to have computer science
without, first and foremost, inventing the computer itself. And it’s impossible to have
a new computer science and have that new field of study
without inventing a new supercomputer that must change the way
we look at the fastest computers. Prior to 1964, computer practitioners
in the United States were not trained as computer scientists. The reason was that back in 1964
the research community had not accumulated enough new knowledge
that would have justified the full time training
of a computer scientist. Just like America must be discovered
before it can be colonized, the supercomputer must be invented
before the computer scientist can be trained. The inventor
of a new supercomputer is the inventor of a new degree
in the new computer science. [How Are Supercomputers Used?] The story of Nigeria encapsulates
the story of humanity. Since 1958,
the economic progress of Nigeria depends of the crude oil and natural gas discovered and
recovered. Only about half of the crude oil discovered
can be recovered. The recovery of crude oil and natural gas
requires a mix of technologies. The parallel supercomputer
is the most advanced technology that is used in the petroleum industry
and used to discover and recover otherwise elusive crude oil
and natural gas that were buried one-mile deep
and that spreads around an area the size of a town. The usefulness of the supercomputer
is the reason one in ten supercomputers are purchased by the petroleum industry. The supercomputer market
is twenty billion dollars per year. Worldwide over two billion computers
are in use and parallel processing
will play a vital role in the development of the computer
of tomorrow. [My Grand Challenge Lectures in Supercomputing] Why are the lecture series
that I posted on YouTube dot com as long as a nice big limousine? My supercomputing lectures
are from the frontiers of knowledge in physics, mathematics, and
computer science. It took me sixty years,
onward of January 1960, to understand how to use the times table that
I learned as a five year-old and do so across a new internet
that is a new global network of millions upon millions
of commodity-off-the-shelf processors that were tightly-coupled to each other
and that shared nothing between each other and to understand
how to use that new knowledge to solve the grand challenge problems arising
at the frontiers of knowledge of extreme-scale mathematics
and computational physics. Parallel processing, the vital technology
that I discovered as underpinning every supercomputer took me sixty years
to fully understand. Therefore, you cannot understand
parallel processing in only sixty minutes and understand
how to solve a grand challenge problem that took me sixty years to understand. A bestselling book
like “Gone with the Wind” will detain you for nearly one thousand
five hundred (1,500) pages. The reason my parallel supercomputer lecture
series that I posted on YouTube
detains you for over one hundred hours is that supercomputing
is far more complex than any novel and that my lecture
on how to parallel process across a new internet is a first person account
of the new knowledge that I accumulated and discovered
across six decades. [Wild applause and cheering for 17 seconds] Insightful and brilliant lecture

One Comment

  1. Philip Emeagwali said:

    i'm Philip Emeagwali. I have three [quote unquote] “first” supercomputers that corresponded to three technologies. The first [quote unquote] “first” supercomputer of 1946 was the first that was programmable. The second [quote unquote] “first” supercomputer of December 1965 was the first supercomputer to be rated at one million instructions per second. I began programming that second [quote unquote] “first” supercomputer on June 20, 1974 in Corvallis, Oregon, United States. The third and most important “first” supercomputer is the precursor to the modern supercomputer of today that can execute large-scale, excruciatingly detailed computational fluid dynamics codes, such as climate models, and execute them across millions of tightly coupled commodity-off-the-shelf processors.

    Serial Processing Versus Parallel Processing Debate

    Back in the 1980s,

    parallel processing was controversial

    and was mocked

    by the community of 25,000

    vector supercomputer scientists.

    Looking back, that widespread

    rejection of parallel processing

    gave it’s acceptance in 1989

    some street cred.

    If you invent something

    that everybody accepted

    then you’ve not invented anything new.

    None of the 25,000 vector supercomputer scientists,

    or their leader, Seymour Cray,

    had a deep understanding

    of practical parallel processing.

    If they understood

    how to solve real world problems,

    as opposed to textbook problems,

    and if they understood

    how to parallel process and how to solve

    grand challenge problems

    and how to solve them across

    an actual ensemble of

    sixty-four binary thousand processors,

    then Seymour Cray

    would not have mocked, ridiculed,

    and dismissed parallel processing

    as a huge waste of everybody's time.

    Seymour Cray

    is best remembered for his famous quote that ridiculed parallel processing

    as the technology

    that will render his supercomputers obsolete.

    Seymour Cray,

    taunted aspiring parallel supercomputer scientists

    by asking them:

    [quote]

    “If you were plowing a field,

    which would you rather use?

    Two strong oxen or 1,024 chickens?”

    [unquote]

    This famous quote of Seymour Cray

    was used to taunt me

    when I was attempting to solve

    the toughest problems

    arising in mathematics and physics

    and attempting to solve them

    by parallel processing them across

    my new internet

    that was a new global network of

    65,536 processors.

    Back in 1989,

    it made the news headlines

    that an African supercomputer wizard

    in the United States

    had figured out

    that 65,536 chickens,

    that were his metaphors

    for his as many commodity-off-the-shelf processors

    that were tightly-coupled to each other,

    and figured out that those processors

    can communicate and compute together and do so as one cohesive whole

    virtual supercomputer

    and do so to become more powerful

    than one strong ox

    that was Seymour Cray’s metaphor

    for the vector supercomputer.

    I’m Philip Emeagwali

    and I’m the African

    supercomputer scientist

    that was in the news back in 1989.

    I was in the news because

    I discovered that

    practical parallel supercomputing

    will become the vital technology

    that underpins every supercomputer.

    Team Versus Individual Contributions

    The comparison of my contributions

    during the 1970s and ‘80s,

    to the collective contributions

    of the other

    25,000 supercomputer scientists

    is an act that lacks historical context

    and perspective, namely,

    my discovery of

    parallel supercomputing

    and comparing that single person's contribution

    to the collective contribution

    of the one thousand employees

    of a supercomputer giant

    that was paid billions of dollars

    to deliver supercomputers

    that embodied my new knowledge

    of parallel supercomputing.

    Mount Everest of Science

    I began programming

    one of the world’s fastest supercomputers

    back on June 20, 1974, at age nineteen,

    in Corvallis, Oregon, United States.

    What is a supercomputer?

    The fastest supercomputer

    occupies the space of a soccer field.

    The fastest supercomputer

    costs the budget of any of the

    forty poorest nations in the world.

    The supercomputer is like the lighthouse

    for science.

    Supercomputing

    at the fastest recorded speed

    is like climbing

    the Mount Everest of science.

    How fast is a supercomputer,

    compared to a computer?

    Supercomputing at the speed of

    one billion trillion

    calculations per second

    is like if each of the world’s

    eight billion persons

    solved 125 billion math problems

    per second.

    If one million

    human supercomputers

    computed at the speed of

    one calculation per second,

    they will take 1.5 billion years

    to complete a calculation

    that takes only one second

    on a parallel supercomputer

    that computed at the speed of

    one trillion billion calculations

    per second.

    My contributions to science, namely,

    practical parallel supercomputing,

    that occurred on the Fourth of July 1989

    and occurred at age thirty-four (34)

    should be measured

    by a different yardstick

    when comparing them

    with my cumulative scientific discoveries

    at age sixty-four (64)

    or later at age ninety-four (94).

    The reason is that in thirty (30) years,

    or sixty (60) years,

    I will have doubled, or tripled, my age

    and developed a larger body of discoveries and inventions.

    Comparing my contributions

    at ages thirty-four (34), sixty-four (64), and ninety-four (94)

    is like comparing the gold medal counts at the Olympic Games

    of different sized nations,

    such as Liberia, Nigeria,

    and the United States.

    In scientific research,

    the number of papers published

    doubles every nine years.

    My quest was for the discovery of practical parallel processing

    that is the vital technology

    that changed the way

    we looked at the supercomputer.

    Contrary to a widely held misconception, an article is not a contribution to science

    per se.

    The fifty million scholarly articles published in year 2010 alone

    are nothing but distant

    background noises.

    Inventing a New Computer Science

    The Grand Challenge problem

    of mathematics and physics

    was solved by inventing

    a new computer science, called

    practical parallel processing.

    It’s impossible to have computer science

    without, first and foremost, inventing the computer itself.

    And it’s impossible to have

    a new computer science

    and have that new field of study

    without inventing a new supercomputer

    that must change the way

    we look at the fastest computers.

    Prior to 1964, computer practitioners

    in the United States were not trained

    as computer scientists.

    The reason was that back in 1964

    the research community had not accumulated enough new knowledge

    that would have justified

    the full time training

    of a computer scientist.

    Just like America must be discovered

    before it can be colonized,

    the supercomputer must be invented

    before the computer scientist

    can be trained.

    The inventor

    of a new supercomputer

    is the inventor of a new degree

    in the new computer science.

    How Are Supercomputers Used?

    The story of Nigeria encapsulates

    the story of humanity.

    Since 1958,

    the economic progress of Nigeria depends of the crude oil and natural gas discovered and recovered.

    Only about half of the crude oil discovered can be recovered.

    The recovery of crude oil and natural gas requires a mix of technologies.

    The parallel supercomputer

    is the most advanced technology

    that is used in the petroleum industry

    and used to discover and recover

    otherwise elusive crude oil

    and natural gas

    that were buried one-mile deep

    and that spreads around an area

    the size of a town.

    The usefulness of the supercomputer

    is the reason one in ten supercomputers are purchased by the petroleum industry.

    The supercomputer market

    is twenty billion dollars per year. Worldwide over two billion computers

    are in use

    and parallel processing

    will play a vital role

    in the development of the computer

    of tomorrow.

    My Grand Challenge Lectures in Supercomputing

    Why are the lecture series

    that I posted on YouTube dot com

    as long as a nice big limousine?

    My supercomputing lectures

    are from the frontiers

    of knowledge in physics, mathematics, and computer science.

    It took me sixty years,

    onward of January 1960,

    to understand how to use the times table that I learned as a five year-old

    and do so across a new internet

    that is a new global network of

    millions upon millions

    of commodity-off-the-shelf processors

    that were tightly-coupled to each other

    and that shared nothing

    between each other and to understand

    how to use that new knowledge

    to solve the grand challenge problems arising at the frontiers of knowledge

    of extreme-scale mathematics

    and computational physics.

    Parallel processing, the vital technology

    that I discovered as underpinning

    every supercomputer took me sixty years

    to fully understand.

    Therefore, you cannot understand

    parallel processing in only sixty minutes

    and understand

    how to solve a grand challenge problem

    that took me sixty years to understand.

    A bestselling book

    like “Gone with the Wind”

    will detain you for nearly one thousand

    five hundred (1,500) pages.

    The reason my parallel supercomputer lecture series

    that I posted on YouTube

    detains you for over one hundred hours is that supercomputing

    is far more complex than any novel

    and that my lecture

    on how to parallel process across

    a new internet is a first person account

    of the new knowledge

    that I accumulated and discovered

    across six decades.

    November 30, 2019
    Reply

Leave a Reply

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