Published December 07, 2025
Day 7 of Advent of Code 2025 has us tracking the splitting of tachyon beams through multiple timelines. For part 1 we care about how often the beam gets split, and in part 2 we care about how many timelines are created. This gives us the nice ability to get the solution for both parts in a single run.
Part 1 was pretty easy, I had it done and working inside of 30 minutes. I went line by line and counted how often the beam encountered a splitter. Part 2 required us to look at every possible split and count them - even if they had already occurred. My first inclination was to create a list of all the points the beams hit on their way down. That might have worked but I did want to actually finish today. I then realized I didn't need to track each one, just how many times it would hit a given point. Once I had that I got a solution that runs in ~600us.
You can check out the full solution on my GitHub.