A Hardware Implementation of the Snappy Compression Algorithm

Kyle Kovacs

EECS Department
University of California, Berkeley
Technical Report No. UCB/EECS-2019-85
May 18, 2019

http://www2.eecs.berkeley.edu/Pubs/TechRpts/2019/EECS-2019-85.pdf

In the exa-scale age of big data, file size reduction via compression is ever more important. This work explores the possibility of using dedicated hardware to accelerate the same general-purpose compression algorithm normally run at the warehouse-scale computer level. A working prototype of the compression accelerator is designed and programmed, then simulated to asses its speed and compression performance. Simulation results show that the hardware accelerator is capable of compressing data up to 100 times faster than software, at the cost of a slightly decreased compression ratio. The prototype also leaves room for future performance improvements, which could improve the accelerator to eliminate this slightly decreased compression ratio.

Advisor: Elad Alon


BibTeX citation:

@mastersthesis{Kovacs:EECS-2019-85,
    Author = {Kovacs, Kyle},
    Title = {A Hardware Implementation of the Snappy Compression Algorithm},
    School = {EECS Department, University of California, Berkeley},
    Year = {2019},
    Month = {May},
    URL = {http://www2.eecs.berkeley.edu/Pubs/TechRpts/2019/EECS-2019-85.html},
    Number = {UCB/EECS-2019-85},
    Abstract = {In the exa-scale age of big data, file size reduction via compression is ever more important. This work explores the possibility of using dedicated hardware to accelerate the same general-purpose compression algorithm normally run at the warehouse-scale computer level. A working prototype of the compression accelerator is designed and programmed, then simulated to asses its speed and compression performance. Simulation results show that the hardware accelerator is capable of compressing data up to 100 times faster than software, at the cost of a slightly decreased compression ratio. The prototype also leaves room for future performance improvements, which could improve the accelerator to eliminate this slightly decreased compression ratio.}
}

EndNote citation:

%0 Thesis
%A Kovacs, Kyle
%T A Hardware Implementation of the Snappy Compression Algorithm
%I EECS Department, University of California, Berkeley
%D 2019
%8 May 18
%@ UCB/EECS-2019-85
%U http://www2.eecs.berkeley.edu/Pubs/TechRpts/2019/EECS-2019-85.html
%F Kovacs:EECS-2019-85