suckerpinch
7 HN recommenders · recommended in 8 months, 2020–2025
19:08 · 314K views · Apr 1, 2019
Tom Murphy (suckerpinch) presents a deliberately trollish computer architecture built on IEEE-754 floating point, using NaN and infinity to define logic gates. He builds working hardware around the idea.
HN fans hold it up as a favorite among his videos and an example of gloriously bonkers technical art, and cite it in floating-point discussions.
9 comments from Hacker News
Seconded, thirded, and fourthed. My favorite is "NaN Gates and Flip FLOPS", where he dismisses 0 and 1 as being too ugly to be the foundation of all computation and reinvents computing via the beauty of rational numbers, which is to say, the horror of IEEE754 floating point (along with a working hardware implementation!): this video
Suckerpinch barely ever releases a video, but when he does they are great. NaN gates and FlipFlops are my favorite: this video You can only get good at mathematics by doing, but if there is a channel that helps you get better at math passively it is probably Michael Penns': [LINK]
Here's my favorite video of his, where he reinvents computing from scratch while eschewing inelegant 1s and 0s for elegant floating-point values: this video
That's a beautiful form of art, somehow. Far away from the only form of art, but it's the kind of art where a) it's very obvious that you have to be extremely proficient in what you are doing (and so the stupid modern art spectator comment "I could have done this" is quelled from its very spark), and b) it's so useless, especially for the amount of effort put in, that it really only exists for itself and to please both creator and spectator. There is one video of his that takes it to an extreme: this video - That is not only impossibly bonkers on just about every level, it's also so incredibly useless that the result is not even observable.
Why stop at half-precision floats? Why not go all the way to 3-bit floats? And build a hardware accelerator out of them? Tom Murphy has you covered with: the NaN Gate. this video [LINK]