high performance computing on graphics processing units: hgpu.org

Key derivation functions are a key element of many cryptographic applications. Password-based key derivation functions are designed specifically to derive cryptographic keys from low-entropy sources (such as passwords or passphrases) and to counter brute-force and dictionary attacks. However, the most widely adopted standard for password-based key derivation, PBKDF2, as implemented in most applications, is highly susceptible to attacks using Graphics Processing Units (GPUs). Due to their highly parallel architecture, GPUs are ideal for performing graphic calculations. In time, it became apparent that GPUs can be also used for wide range of other practical applications (including cryptography). In this work, we analyze how the design of PBKDF2 allows efficient attacks using GPUs and discuss possible alternatives that address this problem. Next, we present and analyze results of PBKDF2 benchmarks run on current CPU and GPU hardware. Finally, we present our demonstration program which utilizes the GPU hardware to perform a brute-force or dictionary attack on a LUKS encrypted partition.