Parallelism divides a workload into tasks that can run across multiple CPU cores. Unlike SIMD, which processes multiple values with one operation on a single core, parallelism uses multiple cores to do work simultaneously. section 12.0
In Unity, Jobs can split data across threads. A data-oriented design with flat arrays provides a foundation for adding parallel processing, though the extra tools increase code complexity. section 12.0 section 12.5
Updating enemy positions
Use Jobs with TransformAccessArray to update enemy GameObject positions in parallel. This adds scheduling and completion-wait overhead, so use it when transform updates are a bottleneck. section 12.5






















































