Today I felt unwell again. I also had a discussion with my father about approximating pi, however, that inspired me to do some research on the subject. I ended up reading the historical section of the Wikipedia article on pi. When I was in college, the consensus was that you should never read about math on Wikipedia. My understanding is that the mathematical articles have improved since then, however, and since I’m sure there are many eyes on this particular article, I decided it would serve my purpose.
As I told my father, ancient mathematicians (such as Archimedes, but also including mathematicians in India and China) calculated pi by approximating the circumference of a circle using the perimeters of polygons with increasingly many sides. According to the article, after 1500, mathematicians in both Europe and India began to use infinite series that could be shown to converge to pi, instead. In the 20th century, they continued to use infinite series, but also developed very fast iterative algorithms in which each step involved applying the same calculation to the results of the previous step. Computers following these algorithms were able to generate millions of digits of pi. Currently, the limiting factor in finding new digits of pi seems to be, not processor speed, but availability of storage for the huge numbers needed for each calculation.