9180

Computing Privacy-Preserving Edit Distance and Smith-Waterman Problems on the GPU Architecture

Shi Pu, Jyh-Charn Liu
Department of Computer Science and Engineering, Texas A&M University, College Station, TX 77840
Cryptology ePrint Archive: Report 2013/204, 2013
@misc{cryptoeprint:2013:204,

   author={Shi Pu, Jyh-Charn Liu},

   title={Computing Privacy-Preserving Edit Distance and Smith-Waterman Problems on the GPU Architecture},

   howpublished={Cryptology ePrint Archive, Report 2013/204},

   year={2013},

   note={url{http://eprint.iacr.org/}}

}

Download Download (PDF)   View View   Source Source   

810

views

This paper presents privacy-preserving, parallel computing algorithms on a graphic processing unit (GPU) architecture to solve the Edit-Distance (ED) and the Smith-Waterman (SW) problems. The ED and SW problems are formulated into dynamic programming (DP) computing problems, which are solved using the Secure Function Evaluation (SFE) to meet privacy protection requirements, based on the semi-honest security model. Major parallelization techniques include mapping of variables to support collision-free parallel memory access, scheduling and mapping of gate garblers on GPU devices to maximize GPU device utilization, and latency minimization of context switch for computing steps in the DP matrix. A pipelined GPU-CPU interface is developed to mask latency of CPU housekeeping components. The new solutions were tested on a Xeon E5504 at 2GHz plus a GTX-680 GPU (as generator), connecting an i7-3770K at 3.5GHz plus a GTX-680 GPU (as evaluator) via local Internet. A 5000×5000 8-bit alphabet ED problem requires roughly 1.88 billion non-free gates, and the running time of around 26 minutes (roughly 1.209×10^6 gate/second). A 60×60 SW problem is computed in around 16.79 seconds. Compared to the state of art performance [5], we achieved the acceleration factor of 12.5x for the ED problem, and 24.7x for the SW problem.
VN:F [1.9.22_1171]
Rating: 0.0/5 (0 votes cast)

* * *

* * *

TwitterAPIExchange Object
(
    [oauth_access_token:TwitterAPIExchange:private] => 301967669-yDz6MrfyJFFsH1DVvrw5Xb9phx2d0DSOFuLehBGh
    [oauth_access_token_secret:TwitterAPIExchange:private] => o29ji3VLVmB6jASMqY8G7QZDCrdFmoTvCDNNUlb7s
    [consumer_key:TwitterAPIExchange:private] => TdQb63pho0ak9VevwMWpEgXAE
    [consumer_secret:TwitterAPIExchange:private] => Uq4rWz7nUnH1y6ab6uQ9xMk0KLcDrmckneEMdlq6G5E0jlQCFx
    [postfields:TwitterAPIExchange:private] => 
    [getfield:TwitterAPIExchange:private] => ?cursor=-1&screen_name=hgpu&skip_status=true&include_user_entities=false
    [oauth:protected] => Array
        (
            [oauth_consumer_key] => TdQb63pho0ak9VevwMWpEgXAE
            [oauth_nonce] => 1474955355
            [oauth_signature_method] => HMAC-SHA1
            [oauth_token] => 301967669-yDz6MrfyJFFsH1DVvrw5Xb9phx2d0DSOFuLehBGh
            [oauth_timestamp] => 1474955355
            [oauth_version] => 1.0
            [cursor] => -1
            [screen_name] => hgpu
            [skip_status] => true
            [include_user_entities] => false
            [oauth_signature] => fZ1wkTzdmG6NLZRF2OFUGXliU08=
        )

    [url] => https://api.twitter.com/1.1/users/show.json
)
Follow us on Facebook
Follow us on Twitter

HGPU group

1998 peoples are following HGPU @twitter

HGPU group © 2010-2016 hgpu.org

All rights belong to the respective authors

Contact us: