Okay, abstracting from the system. What I meant is that bashtop takes in general about 5 times cpu cycles than that of htop. Since htop, top and bashtop report the numbers differently I used ps to check 3 of them:
ps -o pid,user,%cpu,cputimes,nlwp,command ax | grep top
40527 ana 17.1 106 1 bash /usr/bin/bashtop
40569 ana 3.0 19 1 htop
40586 ana 0.3 2 1 top
I really don’t understand why. I checked the source, but with my knowledge of bash is hard to figure it out.