4 ms·
That's true and a nice exercise for a programming class, but if you "reduce the problem to fixed stride for loops over arrays that can be fully pipelined." to m
by Someone 10y ago
That's true and a nice exercise for a programming class, but if you "reduce the problem to fixed stride for loops over arrays that can be fully pipelined." to make the game of life as fast as possible, you're doing it wrong (for those who don't know about it: https://en.wikipedia.org/wiki/Hashlife https://en.wikipedia.org/wiki/Hashlife)
- nimos 10y agoYeah for sure. We only had to calculate 1 or 2 iterations of the game and had a random seeded array as input and preallocated array we had to populate as output so there was a lot of optimization that wasn't worth it if you were actually looking to implement Game of Life, it was designed specifically to get you to think about computer system stuff.
- tomovo 10y agoIt's also true and nice for most modern game engines. I'm quite sure Game of life was just a subject for the exercise with the purpose to show that you can get major speedups without changing the algorithm - you don't always have a better one as backup.
- adrianratnapala 10y agoHashlife is impressive, but don't misread the above comment as saying it is "the" right way of doing it. Hashlife is good if there is lots of repetition (or sparseness). There are many other optimisation strategies, with different trade-offs.