3 ms·
Great idea! I don't know the optimal solutions, only that each task can be solved. Perhaps someone smarter than me can prove that a certain solution is optimal.
by sergex 8y ago
Great idea! I don't know the optimal solutions, only that each task can be solved. Perhaps someone smarter than me can prove that a certain solution is optimal. But I could record the solutions provided and show if someone else have solved the task with fewer components or less nand-gates.
It would also be cool to show the overall cost in nand-gates or even transistors. And to be able to show the solutions decomposed into only transistors and wires.
Didn't know Euclid the Game, but it seems very cool and close to what I wanted to achieve.
- n4r9 8y agoI'd be happy to have a crack at the optimality part. The first few are obvious but it gets quite interesting. Email my username at protonmail.com if you're interested =]
- meta_AU 8y agoIf you have a look at any Zachtronics games you can see one way of doing it. It is just a leaderboard represented as a histogram. A separate one for instructions, cycles, etc. No need to know the best, just keep track of what other people have achieved.
- jakear 8y agoGate count is interesting from a "how much would this const to make" perspective, but I think minimum clock cycle time is a more interesting metric, as that will show how fast you can run the thing. Maybe a two dimensional plot? It would be very interesting to have a worldwide "leaderboards" section with a dot on the plot for every entry, and a link to see how they did it.
- sergex 8y agoInteresting idea, but how should this be determined? Would it be a "cost" per transistor and wirelength, and then the minimum clock cycle time would be the "worst case" path through the circuit?
- jakear 8y agoYes exactly. The 6.004 course at MIT already has software that does this sort of thing.
- sergex 8y agoThanks, it is a very cool idea. I wanted to make the game really simple and accessible, but there seem to be interest in optimization challenges like finding the smallest/cheapest/fastest possible solution.