2 ms·
Not only that, but the O(log n) sparse table has multiple branch points that are more predictable, so the branch predictor works better. It can be faster overal
by naasking 9y ago
Not only that, but the O(log n) sparse table has multiple branch points that are more predictable, so the branch predictor works better. It can be faster overall than a single highly unpredictable branch point.