I’m Philip Emeagwali. I invented a new supercomputer that could be used to solve the toughest problems arising in mathematics and physics. Such problems are called the Grand Challenge Problems of supercomputing. Back in 1989 the year I completed my invention and began appearing in major U.S. newspapers, Seymour Cray —who was the then leading mind in the world of vector processing supercomputers—could not understand the mathematics of where each of my 64 binary thousand initial-boundary value problems should be and when it should be

at each of my as many processors.

Back in 1989,

the supercomputer scientists

that attempted to parallel process across an ensemble of processors

were processing

by the seat of their pants.

That is, those supercomputer scientists of the 1970s and ‘80s

did not understand

the complicated mathematics

and did not have the command

of scientific materials

and the subject matter knowledge

that was needed

to solve the Grand Challenge Problem. The Grand Challenge Problem

was at the crossroad

where calculus, algebra, physics,

computing, and supercomputing

met each other.

Because they did not understand

the Grand Challenge Problem,

the 25,000 vector processing supercomputer scientists of the 1980s and earlier

began to hate

the massively parallel processing supercomputer

that I used

to solve the Grand Challenge Problem.

Back in the 1980s, I was the only

internet scientist

that knew the 64 binary thousand,

or the 65,536, “address tags”

that directed where on my internet

each of my 65,536

initial-boundary value problems

of calculus and physics

was delivered.

To the 25,000 vector processing supercomputer scientists of the 1980s,

reading my 65,536

email message-passing codes

was as incomprehensible

as reading a Chinese newspaper.

My invention

of how to provide the address tags

was the necessary pre-condition

to the invention

of the massively parallel processing supercomputer

that is also a new internet de facto.

That invention

has rich and fertile consequences

and contributed

to the more complete understanding

of how and why

the technology called

parallel processing—or solving

millions upon millions of problems—

across as many processors

and solving them at the same time

makes the computer faster

and makes the supercomputer super.

19.1.2 Birth of a New Computer

My quest for the precursor

to the modern supercomputer

that is fastest by parallel processing across

a new internet

that is a new global network of

processors

began as a vague idea.

That quest began

as the seed of an Iroko tree

and blossomed, sixteen years later,

into the world’s fastest supercomputer

that is the Iroko tree

of the unknown forest

named the massively parallel processing

supercomputer.

The Iroko tree

is the tallest tree in Igbo Land

of southeastern Nigeria.

The Iroko tree

grows along the west coast of Africa.

The Iroko tree

can live for up to 500 years.

My quest for the parallel processing supercomputer

began in the early morning of

Thursday June 20, 1974

in Corvallis, Oregon, United States.

My quest for the modern supercomputer

ended at 8:15 in the morning

of Tuesday the Fourth of July 1989

in Los Alamos, New Mexico, United States.

That quest for the parallel processing

supercomputer

led to my deeper and surer

understanding of the Internet

as a planetary supercomputer-hopeful.

I had ideas about parallel processing

since 1974.

But, until the Fourth of July 1989,

I did not experimentally prove

that parallel processing

makes the impossible-to-compute

possible-to-compute.

My experimental discovery

of parallel processing

that occurred across my new internet

that is a new global network of

65,536 processors

occurred on the Fourth of July 1989.

Philip Emeagwali Lecture 180913-2