nostr:npub1eyyf4k2tdwv4lrnmec82wdjlf8v377p6hjew8lrr0yw7sxqfcy7qh3nw7n From what I now about how CPUs execute instructions, this can only be a rough estimate. But very handy indeed!

From looking at what you are doing, have you considered using #SIMD instructions? 

The term seemed scary to me at first, but it's really not that complex. The performance benefit is huge, even on older CPUs. Here's a random example code to give you an idea: https://github.com/jean553/c-simd-avx2-example