Fun with Numbers: The Pan-Prime-Digit Cube Hypothesis

[DavidM]

(08-23-2017 11:53 PM)Claudio L. Wrote:  OK, so here are a couple of thoughts:
I think the optimal code works by adding digits to the number one by one (without looking at his sources, I guarantee you this is Werner's method, or very similar).

Given that I had already implemented a couple versions of this for the PC (using a Delphi app) -and- having already seen Werner's description of his method, you have almost exactly described the method I was already planning to try on the 50g.

And, yes, 6 digit suffixes should number 4880. 7-digits yield 19520, 8-digits yields 78176, etc.

In order to speed things up a bit I'm planning to pre-build a list of 4880 6-digit ZINTS as a placeholder for all the suffixes from each "round". With each additional digit prefix, I'll update the zints and their count (in place via a Saturn routine) for the next round. That will keep me from having to create new suffix lists on the fly, and should reduce some of the GC events as well. Once I've reached the 6-digit suffix realm, I'll just keep it at that and run successive tests with incremental values of (n*10^6)+suffix.