Each thread now handles a single word at a time instead of a whole list. This allows the workload to be more evenly distributed across threads and stay more parralel
3.5 KiB
3.5 KiB