Academia is the best place to do novel research, and a bad place to write great codes (like Tarsnap). You can't do great research on "Birch and Swinnerton-Dyer

Academia is the best place to do novel research, and a bad place to write great codes (like Tarsnap). You can't do great research on "Birch and Swinnerton-Dyer conjecture", if you are not an academian (like Manjul Bhargava). > In 2005, I made the first publication of the use of shared caches in multi-threaded CPUs as a cryptographic side channel... As far as I know, the first publication about cache side channel attacks was [1] by Dan Page in 2002. [1] https://eprint.iacr.org/2002/169.pdf https://eprint.iacr.org/2002/169.pdf