Analysis of Results

First of all using this equation requires a significant number of terms to compute the value of $\pi$, at least 3092300000 terms in this series. By adjusting the block size or the chunk of work the amount of communication is reduced an the efficiency stays acceptable through using 8 nodes with the large chunk size, with a block size of 1000000 or greater. Also The efficiency is marginal using any chosen block size at 1000000 or above and 16 or 24 nodes. With a small block size of 100000 the application can only effectively use 4 processors.