This is a parallel Pohlig-Hellman + baby-step giant-step implementation made for the final project of Parallel and distributed systems, paradigms and models course at unipi.
It employs a few tricks to make computation fast and somewhat scalable, namely:
- all the standard implementation tricks of sequential bsgs,
- shared hash table exploiting atomic operations,
- implicit storage in the table of the resulting element
g^i, avoiding allocating big nums and improving memory usage.