3 ms·
You still need the same amount of iterations (79 685 186 856 218)[1] to resolve the problem but it's now faster because SIMD instructions[2] lowered the amount
by CodeArtisan 7y ago
You still need the same amount of iterations (79 685 186 856 218)[1] to resolve the problem but it's now faster because
SIMD instructions[2] lowered the amount of cpu cycles needed per operation.
Novel algorithms[3] lowered the amount of operations needed per iteration.
[1] https://people.csail.mit.edu/rivest/lcs35-puzzle-description.txt https://people.csail.mit.edu/rivest/lcs35-puzzle-description... (short description at the bottom).
[2] https://en.wikipedia.org/wiki/SIMD https://en.wikipedia.org/wiki/SIMD
[3] https://gmplib.org/manual/Algorithms.html https://gmplib.org/manual/Algorithms.html
- vlovich123 7y agoThe problem appears to be sequential so how would SIMD help? Also what advances in squatting algorithms have there been specifically? None of the links you have provided appear relevant
- eru 7y agoThe numbers involved are quite large. They don't fit into a single computer word. The problem is designed so that iterations have to be sequential, but a single iteration can use parallelism.