Buchstab's function is a tool in number theory that measures how common "rough numbers" are. A rough number is a positive integer that has no small prime factors, meaning all of its prime factors are at least a certain size. For example, if you want to count how many integers up to a billion have no prime factors smaller than 1000, Buchstab's function tells you roughly what fraction of integers qualify. This function is important in analytic number theory, appearing in problems related to prime gaps, sieve methods, and the distribution of primes.
The challenge with Buchstab's function is that it is defined through a delay differential equation, meaning its value at any given point depends on its values at earlier points. While the function can be computed numerically by solving this equation step by step, that process is cumbersome and not always practical when you just need a quick, reliable estimate. Previous approximations existed but were either hard to work with or lacked rigorous error guarantees.
The authors of this paper derive explicit, fully rigorous upper and lower bounds for Buchstab's function that are simple formulas you can just plug numbers into directly, without solving any differential equations. The bounds are tight enough to be genuinely useful in mathematical proofs and computations. This kind of explicit, verifiable bound is especially valuable in computational number theory, where researchers need guaranteed error margins rather than just approximate numerical answers.