Hi Bernard. Can you shed some light on why Rivest's prediction that this puzzle would take 35 years of computation was so wildly off? In the paper he predicts that clock speeds would reach 10Ghz by 2012, he also predicts another 5 fold speed improvement by 2034. I am not sure what the actual technological improvements have been since 1999 but it seems our processors are way below his predictions of where we would be in 2012, let alone 2034. He states:

> "We estimate that the puzzle will require 35 years of continuous computation to solve, with the computer being replaced every year by the next fastest model available. Most of the work will really be done in the last few years, however."

So how were you able to perform a calculation that should take 35 years in just 4? Where did Rivest get his assumptions wrong?