Moore's Law

Learning target: students can make predictions based on exponential growth

Right off the bat I throw this packet at them and ask them to read it.  Its the MIT photo essay of various computer chips and transistors throughout computer history.


I ask students what their questions are and I get a range of answers including:

"What is this?"
"How does this relate to what we are doing in class?"
"Why are we reading this?"

So I tell them that the first picture is the birth of the digital age, and give them a brief overview of what a transistor is and how it related to computing power.

From their questions it was clear I was not going to get a question that would involve making a prediction so I gave them the task:

1. how many transistors are their in a computer chip made in 2013/2014? 

(This we can look afterwards and see what reasonable answer would be)

2. When will the number of transistors pass the trillion transistor mark?

Step 1: make a guess.  I ask students to make a guess, based on what their gut says and the data points. I can steer them a little "what number do you know it has to be bigger than?"  Studies show student engagement increases if they make a guess 1st.  I had students write down their guesses and I ut them into an excel file.

Step 2: attack!  It took a while for students to even begin to start this problem.  Then eventually someone figured out they had better make a table just to get something.

Then they found the common differences and realized that wasn't useful, so to set students up for the next part I told them a graph was another strategy they could try.  After some time I got the question "Mr. Olson, what should the scale be?  How am I going to fit 1 and 371 million on the same graph?"  -this will be a great lead in to logarithms.

While trying to make an exponential line of best fit students needed to know how to find the average ratio from year to year.  This proved difficult since the data isn't pretty and the years aren't uniform.