AMD Ryzen is a CPU. I wonder if lightmass can’t parallelize enough to use all the cores? I think I read somewhere that each lightmass performance is limited to the largest mesh so if you only have a few huge meshes you wont be able to parallelize the workload as well
I found where I read it, this is from the docs:
“Optimize the lightmap resolutions across the map so that build time for meshes is more even. The lighting build can never be faster than the slowest single object, regardless of how many machines are doing the distributed build. Avoid large continuous meshes that surround a large part of the level and with high lightmap resolution. You will get better build times if you break these up into more modular pieces, especially on machines with many cores.”