 So, this example has two tasks, and we'll be using machine A as the baseline again. Machine A, the moment spends one-third of its time processing task one, and the other two-thirds of its time processing task two. On the other hand, machine B is much better at running task number two, and can do that in half as much time. So, looking to see how much faster machine B is than machine A, is machine B is obviously going to require less time. So again, we'll begin by calculating how much time each of these machines should take to run. Again, machine A is spending one-third of its time on task number one, and it's spending the other two-thirds of its time on task two. So obviously, that's 100% of its time. It's spending all of its time on this workload at the moment. Machine B does just as well at running task one. So it's going to spend this same amount of time running task one. It's not really important exactly how much time this is, just that we can quantify it as this one-third of however much time machine A spent before. Now, machine A spent the other two-thirds of its time on task two, but machine B is better at this task. So we expect that it's going to take less time, how much less, half as much. So, pulling the one-half, and then we have the same amount of work that we had before. So I have one-third of this time, plus half of two-thirds, which is one-third. Add those together, and I get a total of two-thirds of this amount of time. And it took machine one to run the same amount of work. So again, I can put this into my relative performance equation. And since I'm interested in to see how much faster machine B is than machine A, I'm going to put machine B on the bottom since I have execution times. So I have one times time on the top for machine A, and two-thirds times T on the bottom for machine B. Obviously, when I reduce this, we'll get three halves. So machine B is 50% faster than machine A. Machine B can get twice as much work done on task two, the same amount of time that machine A spends on that task.