Slowhashing

Slowhashing refers to cryptographic techniques that deliberately increase the time, memory, and computational resources required to compute a hash of a password or other secret. The goal is to resist offline brute-force and dictionary attacks by making each guess expensive, so that attackers cannot test large numbers of guesses quickly.

Implementations typically use memory-hard or time-costly algorithms, with tunable parameters such as iteration count, memory usage, and parallelism. A per-input salt is used to prevent precomputation and rainbow tables. Proper parameters balance security with acceptable authentication latency on legitimate systems.

Common slowhashing algorithms include Argon2 (with variants Argon2i, Argon2d, Argon2id), scrypt, bcrypt, and PBKDF2. Argon2 and scrypt are memory-hard, while bcrypt and PBKDF2 rely more on CPU time; all support adjustable costs. Some algorithms emphasize memory access patterns that are hard to parallelize on GPUs.

Applications are primarily in password storage and key derivation for encrypted data or systems. When used for password hashing, slowhashing should be combined with unique per-password salts and, where appropriate, peppers. Inadequate parameters or implementation flaws can undermine security or create performance problems.

See also: memory-hard function, password hashing, key derivation function, Argon2, scrypt, bcrypt, PBKDF2.