How can one identify the part of the program that is the bottleneck and slows down the overall system performance?

Profiling is used to identify bottlenecks in a program — measuring the execution time of different parts of the code.

Main methods:

• Profilers: tools like gprof, Valgrind (Callgrind), perf allow collecting statistics on functions.
• Timing measurements: inserting timers (for example, using std::chrono in C++) around suspicious sections.
• Resource analysis: monitoring CPU, memory, input/output usage.

Example using std::chrono:

#include <iostream>
#include <chrono>

void functionToTest() {
    // code
}

int main() {
    auto start = std::chrono::high_resolution_clock::now();
    functionToTest();
    auto end = std::chrono::high_resolution_clock::now();
    std::chrono::duration<double> diff = end - start;
    std::cout << "Execution time: " << diff.count() << " seconds\n";
}

Thus, the most time-consuming sections are identified and optimized.