Argon2id settings

[grb]

@Mulled7768

the consistent advice in the RFC is to set parallelism at twice the number of cores (See section “Parameter Choice”) and to 4 where the cores are not unspecified (presumed to be two). Why on earth would they do this if reducing p improved security?

This is my take: Increasing parallelism makes it possible for the user to use higher memory settings while maintaining an acceptable unlock/login delay. If (and only if) the memory is increased in direct proportion to the parallelism increase, there will be a net disadvantage to a GPU-based attacker. Therefore, if your device has surplus memory resources, it makes sense to maximize parallelism.

I have tried to explain this in more detail in an older thread:

Argon2id Parallelism explanation

Again, my understanding of this is not 100%, so I may have some of the details incorrect, but this is how I think about it: Let’s consider an RTX 4090 GPU, which has 16,384 cores and 24 GB of memory shared between all cores. No matter what KDF settings you have, the attacker can divide and conquer the brute-force search space to speed up the search (e.g., letting one core test all passwords that start with A, while the next core tests for all passwords that start with B, etc.). With PBKDF2 as…