GMP used during 3 and a half years to solve MIT's LCS35

Bernard Fabrot bfabrot at
Thu May 2 17:45:47 UTC 2019

Re everybody,

I should be talking at the MIT's CSAIL on the 15th of May during the "time
capsule" opening ceremony, very probably showing slides and if so I'll of
course talk about GMP.

Do any of you have any idea how I could relatively easily find how many
modular exponentiation GMP was capable of in 1999 on top of the line
consumer hardware back then?  The problem was calibrated with 3000
squarings and modulo per second back then with n being a 616 digits / 2048
bits number.

One way would be to just find a working machine from 1999 and install a
version of GMP from back then and test it: I don't know if I'll be able to
do that from here till the 15th of May. Is the code from GMP from 1999
easily available? Was it already shipping with Linux distro back then?

I hope I'm not bothering you guys too much with these emails: this LCS35
thing is a pretty epic story to me : )

Here's the link to the original problem description where the parameter of
"3000" can be seen (I don't know exactly how it was chosen):

More information about the gmp-discuss mailing list