Voxel Space[1] could have used this, would that multicore had been prevalent at the time. I recall being fascinated that simply facing the camera north or south would knock off 2fps from an already slow frame rate.

Many of our maps' routes would be laid out in a predominately east or west-facing track to max out our staying within cache lines as we marched our rays up the screen.

So, we needed as much main memory bandwidth as we could get. I remember experimenting with cache line warming to try to keep the memory controllers saturated with work with measurable success. But it would have been difficult in Voxel Space to predict which lines to warm (and when), so nothing came of it.

Tailslayer would have given us an edge by just splitting up the scene with multiprocessing and with a lot more RAM usage and without any other code. Alas, hardware like that was like 15 years in the future. Le sigh.

[1] https://en.wikipedia.org/wiki/Voxel_Space