Instructions compile, run, it will be slow change nb iterations to 20 run with valgrind --tool=callgrind look at output with kcachegrind change fibo call to fibo2 observe the change in kcachegrind