Institutional Repository [SANDBOX]
Technical University of Crete
EN  |  EL

Search

Browse

My Space

Acceleration of the Adiantum algorithm on embedded GPUs

Georgis Polykratis

Full record


URI: http://purl.tuc.gr/dl/dias/E078AE1A-6F75-4A97-9C5E-1D577F47C1D5
Year 2022
Type of Item Diploma Work
License
Details
Bibliographic Citation Polykratis Georgis, "Acceleration of the Adiantum algorithm on embedded GPUs ", Diploma Work, School of Electrical and Computer Engineering, Technical University of Crete, Chania, Greece, 2022 https://doi.org/10.26233/heallink.tuc.92571
Appears in Collections
Relations with other Items

Summary

In the age of information the need for security becomes more and more obvious. Cryptography is a powerful tool providing the necessary confidentiality on sensitive data. There are many cryptography algorithms, each with its own strengths and weaknesses that constantly compete against numerous threats and of course, time. Adiantum is a lightweight alternative providing, for the time being, both speed and security. This thesis focuses on speeding up of this algorithm on large plaintexts using the parallelism offered by a modern GPU. It consists of the detailed profiling and analysisof the algorithm followed by the implementation of its slowest component,XChacha Stream cipher, on Jetson Xavier NX developer kit. Results suggesta x6 speedup on Adiantum_XChacha20, x4 on Adiantum_XChacha12 andx3 Adiantum_XChacha8 with a x4.5, x3.5 and x2.5 energy reduction on thethree versions respectively. These results lead to the conclusion that, usingthe GPU, it is possible to upgrade security by using Chacha20 without compromising neither, speed or energy consumption.

Available Files

Services

Statistics