>"The key insight here is that we can use this technique to delay carry propagation until the end. We can’t avoid carry propagation altogether, but we can avoid it temporarily. If we save up the carries that occur during the intermediate additions, we can propagate them all in one go at the end."

Brilliant and tremendously interesting!

Useful for if/when I write a fast super-long integer math library in the future (yes, I know GNU already has one, but it's always interesting to know/understand tricks like this for >64-bit, AKA "too-long-to-fit-in-a-single-register" integers...)