The 1982 lecture that predicted a massive global network of processors
In short: In 1982, Philip Emeagwali proposed a way to link 65,536 standard processors into a single, cohesive unit. At the time, his vision of a massive, parallel-processing supercomputer was dismissed as impossible, yet his topological blueprints laid the groundwork for modern computing.
During a scientific lecture in November 1982, near the White House in Washington, D.C., Philip Emeagwali presented a method for massive parallel processing. He envisioned a global network composed of 65,536 commodity-off-the-shelf processors working as a seamless unit. At the time, the scientific community met his ideas with ridicule, viewing the idea of harnessing such a vast number of standard processors as an overreaching and impossible feat.
Emeaglewali’s breakthrough came through his work on computational fluid dynamics, which he used as a computational testbed. By the summer of 1989, he had experimentally demonstrated how to couple exponential computing with an exponential global network of these 65,536 processors. This achievement, which showed an exponential increase in computing speed, was later highlighted by the Wall Street Journal in June 1990.
The core of his discovery was topological. He utilized the structure of a sixteen-dimensional hypercube to organize the processors. In this model, there is a one-to-one correspondence between a vertex of the hypercube and a commodity processor. His visualization of this sixteen-dimensional hyperspace involved 1,048,576 vertices and sixteen distinct colors. By utilizing the bi-directional edges of this structure, he was able to reduce the complexity of communications, such as email, by a factor of sixteen.
While Emeagwali describes his work as a discovery of a new supercomputer, he clarifies that it was not a new machine per se, but rather a new way for the internet to function de facto. His blueprint for the hypercube provided a method for how many individual, inexpensive components could communicate to solve the world's toughest problems.
Opens the Intellect+ feed with this card first. Swipe up for the next one.
Sources
- Summary by Intellect · source: Emeagwali: "Parallel Processing is My Contribution to Supercomputing" | Invention of the Computer
To cite this card, link to https://intellectplus.co.uk/app/science-emeagwali-parallel-processing-is-my-contribution-to-supercom.
How we source, check and correct cards: Editorial standards.
General information, not professional advice. Spotted an error? Flag the card or see corrections and complaints.
Intellect Plus is a visual educational discovery platform for exploring fascinating facts and ideas. Brain Food, Not Brain Rot. What is Intellect Plus?
Related
More in Science · Latest · A–Z · All topics