8891

An Implementation of Conflict-Free Offline Permutation on the GPU

Akihiko Kasagi, Koji Nakano, Yasuaki Ito
Department of Information Engineering, Hiroshima University
Third International Conference on Networking and Computing (ICNC), 2012
@inproceedings{kasagi2012implementation,

   title={An Implementation of Conflict-Free Offline Permutation on the GPU},

   author={Kasagi, A. and Nakano, K. and Ito, Y.},

   booktitle={Networking and Computing (ICNC), 2012 Third International Conference on},

   pages={226–232},

   year={2012},

   organization={IEEE}

}

Download Download (PDF)   View View   Source Source   

1346

views

The Discrete Memory Machine (DMM) is a theoretical parallel computing model that captures the essence of the shared memory access of GPUs. The bank conflicts should be avoided for maximizing the bandwidth of the shared memory access. Offline permutation of an array is a task to copy all elements in array $a$ into array $b$ along a permutation given in advance. The main goal of this paper is to implement a conflict-free permutation algorithm on the DMM in a GPU. We have also implemented straightforward permutation algorithms on the GPU. The experimental results for 1024 float numbers on NVIDIA GeForce GTX-680 show that a straightforward permutation algorithm takes 246ns and 877ns for random permutation and bit-reversal permutation, respectively. Quite surprisingly, our conflict-free permutation algorithm runs in 165ns both for random permutation and for bit-reversal permutation although it performs more memory access operations. It follows that our conflict-free permutation is 1.5 times faster for random permutation and 5.3 times faster for bit-reversal permutation.
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] => 1475281361
            [oauth_signature_method] => HMAC-SHA1
            [oauth_token] => 301967669-yDz6MrfyJFFsH1DVvrw5Xb9phx2d0DSOFuLehBGh
            [oauth_timestamp] => 1475281361
            [oauth_version] => 1.0
            [cursor] => -1
            [screen_name] => hgpu
            [skip_status] => true
            [include_user_entities] => false
            [oauth_signature] => PP0fyGgoyylbSp3YCVX9WbNLK4E=
        )

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

HGPU group

2005 peoples are following HGPU @twitter

HGPU group © 2010-2016 hgpu.org

All rights belong to the respective authors

Contact us: