This document is what created the National Center for Supercomputing Applications , among other things:
It’s a good read, and it’s compelling that it was successful in garnering investment. I feel like a lot of the potential has gone unfulfilled, and a generation or so later, I feel like we’ve lost a lot of ground in terms of training programmers to make the most out of these supercomputer-style architectures (which are now part of most consumer computing devices).
Reading this has inspired me to evaluate what a “sequel” to this proposal might look like today. Something that would inspire the current and future generations of programmers to lay down the hoe and learn how to realize the true potential of modern computing machinery, and turn those results over to not only University scientists, but to everyone who has an idea that can be explored more effectively through high-performance computing.