Multi-core, region-aware chunk loading for busy Forge servers.
When many players explore, vanilla tends to pile chunk work onto a single thread. QuantumChunks routes chunk preparation and parts of generation onto a smart worker pool, parallelizing across CPU cores while serializing work inside the same 32×32 region to keep worlds safe and stable.
Minecraft often asks for a “background executor” to run chunk tasks.
QuantumChunks swaps those calls to a dedicated worker pool.
Each task is tagged with its 32×32 chunk region:
Same region → processed one at a time (no races)
Different regions → processed in parallel across your cores
Result: smoother TPS and fewer MSPT spikes when many players move at once.
config/quantumchunks-server.toml
Use your favorite profiler (e.g., Spark):
/quantumchunks - CPU average usage, chunks loaded for 10 secs/1 minute
/spark tps — watch MSPT during exploration
/spark profiler --timeout 60 — look for chunk work on worker threads
You should see higher total CPU usage (multiple cores active) and fewer >50ms ticks during chunk storms.
