the hardest part of the optimization process is … affects the performance of the application. … the processor [can] tell you exactly where and … [via the] on-chip performance monitoring (PM) hardware found on … [p. 58] the Performance Application Programming Interface (PAPI) library … is intended to be completely portable … performance tools and instrumentation that work on one … PAPI supports measurements per-thread
Google-1
Google-2
Copyright clearance needed for quotation.