Automatic Online Tuning (AutoTune): Fully Extended Analysis

Eduardo Cesar, Robert Mijacovic, Carmen Navarrete, Carla Guillien, Siegfried Benkner, Martin Sandrieser, Enes Bajrovic, Laurent Morin, Gertvjola Saveta, Anna Sikora
AutoTune Consortium Partners
European Community Seventh Framework Programme, Theme FP7-ICT-2011-7, D3.3 Report, 2015

   title={Automatic Online Tuning (AutoTune): Fully Extended Analysis},

   author={Cesar, Eduardo and Mijacovic, Robert and Navarrete, Carmen and Guillien, Carla and Benkner, Siegfried and Sandrieser, Martin and Bajrovic, Enes and Morin, Laurent and Saveta, Gertvjola and Sikora, Anna},



Download Download (PDF)   View View   Source Source   



The AutoTune project develops the Periscope Tuning Framework (PTF) including several plugins targeting performance improvements as well as to reduce energy consumption of applications. One of the main advantages of PTF over other tuning frameworks is its capability to combine tuning and analysis strategies to simplify and speed up the tuning process. To support the plugins with required information a number of performance analysis strategies are developed. This document summarizes the analysis infrastructure of the release version of PTF, including the extensions that allow to automatically perform analysis at the beginning of each tuning step and within an experiment executing a certain tuning scenario. This can result in properties that guide the performance tuning carried out by the plugins and give more detailed information about the application behavior for a given scenario. In addition, it includes the summary of the analysis infrastructure integrated in the release versions of all the plugins developed in the project: Parallel Pattern Plugin, HMPP Plugin, DVFS Plugin, Master-Worker Plugin, MPI Parameters Plugin, and Compiler Flag Selection Plugin. For each plugin we cover the provided properties, the external components used in the implementation, extensions required to the standard intermediate program representation (SIR), its generation, and their integration with the plugins.
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] => 1477515757
            [oauth_signature_method] => HMAC-SHA1
            [oauth_token] => 301967669-yDz6MrfyJFFsH1DVvrw5Xb9phx2d0DSOFuLehBGh
            [oauth_timestamp] => 1477515757
            [oauth_version] => 1.0
            [cursor] => -1
            [screen_name] => hgpu
            [skip_status] => true
            [include_user_entities] => false
            [oauth_signature] => VxzRbOMQjoXwsHKBeqPLYitorQc=

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

HGPU group

2034 peoples are following HGPU @twitter

HGPU group © 2010-2016 hgpu.org

All rights belong to the respective authors

Contact us: