A CPU is a general-purpose processor that executes sequential and parallel computations. Modern CPUs commonly provide multiple cores, but using them does not automatically improve performance. section 7.2.2
For a practical test, parallelize a bootstrap confidence-interval function with Python's ThreadPoolExecutor, then compare execution times rather than trusting the generated code's claims. section 7.2.2
When more CPU threads make code slower
A four-thread bootstrap implementation distributes resampling work across CPU threads, but the measured result can be slower than the optimized single-thread version. Low processor utilization and Python's Global Interpreter Lock can limit gains, so benchmark the actual workload. section 7.2.2
