https://www.csail.mit.edu/news/programmers-solve-mits-20-year-old-cryptographic-puzzle
https://blockchair.com/bitcoin/block/573138
Twenty years ago MIT introduced a puzzle that was specifically designed to not be parallelizable, predicting that it wouldn't be solvable in 35 years.
If I'm getting it right it involves computing a square root of a number a fixed number of times. There is no shortcuts to this, and it it not parallelizable: you need the previous root to compute the next one. (I also suppose that it is easy to verify the solution)
As such it makes for a neat Proof of Work algorithm. Even if it doens't fit the PoW Bitcoin currently has (need to be able to distribute the work), I thought that this was a neat piece of news, and some fresh air in current events.
No comments:
Post a Comment