Primerjava izvedb kriptografskih algoritmov na CPE in GPE

ZUPAN, KRISTIAN (author), Dobravec, Tomaž (mentor)

V magistrskem delu smo se ukvarjali s primerjavo zaporednih implementacij bločnih šifer za CPE in njihovimi vzporednimi implementacijami za GPE. Pri tem smo si izbrali vseh pet finalistov standarda AES (Rijndael, Serpent, Twofish, MARS in RC6). Algoritme smo analizirali ter preučili možne izboljšave za njihovo vzporedno implementacijo. Implementirali smo vzporedne implementacije z razporejanjem podatkov ter merili pohitritev v primerjavi z zaporednimi implementacijami ter dosegli do dvajsetkratno pohitritev nekaterih algoritmov. Naredili smo tudi primerjavo med CUDO in OpenCL, platformama za pisanje večnitnih programov za grafične kartice. Implementirali smo tudi popolnoma svoje implementacije z bitnimi rezinami algoritmov Rijndael in Serpent za platformo CUDA ter ju primerjali z vzporednimi implementacijami z razporejanjem podatkov.

• Language: Slovenian
• Keywords: vzporedni algoritmi, bločne šifre, kriptografija, CUDA, OpenCL, bitne rezine, AES, CTR, hitro šifriranje
• Work type: Master's thesis/paper
• Organization: FRI - Faculty of Computer and Information Science
• Year: 2015
• PID: 20.500.12556/RUL-30833
• Publication date in RUL: 19.06.2015

Secondary language

• Language: English
• Title: Comparison of implementations of cryptographic algorithms on the CPU and GPU

Abstract

The aim of this Master's Thesis was to compare the serial implementations of block ciphers that run on CPU with corresponding parallel implementations that run on GPU. By analyzing the five finalists of the AES competition (Rijndael, Serpent, Twofish, MARS and RC6) we searched for possible improvements in their parallel implementations. Using the data parallelism techniques we implemented the algorithms in parallel and achieved the speed that was 20 times higher in comparison to the underlying serial implementations. We have also compared two different platforms for writing parallel programs on GPU: CUDA and OpenCL. In addition we implemented the bitslice implementations of algorithms Rijndael and Serpent for CUDA platform and compared them to data parallelism based implementations.

• Keywords: parallel algorithms, block cyphers, cryptography, CUDA, OpenCL, bitslice, AES, CTR, fast encryption