I'm still kicking myself because when I first heard about Bitcoin (in 2010) I was working on a research project that required three GPU clusters (each costing between $15k and $20k). The contract for the project was delayed and I seriously thought about "benchmarking" the hardware by Bitcoin mining until work on the project could continue. I never actually did any mining and the hardware sat idle in our lab for a couple of months. I don't have the stomach to calculate what the Bitcoins that hardware was capable of generating would be worth today.