search query: @keyword rinnakkaislaskenta / total: 27
reference: 4 / 27
« previous | next »
Author:Piispanen, Matias
Title:Simulating timing and energy consumption of accelerated processing
Kiihdytetyn laskennan ajoituksen ja energiankulutuksen simulointi
Publication type:Master's thesis
Publication year:2014
Pages:73      Language:   eng
Department/School:Perustieteiden korkeakoulu
Main subject:Ohjelmistotekniikka   (T3001)
Supervisor:Saikkonen, Heikki
Instructor:Hirvisalo, Vesa
Electronic version URL: http://urn.fi/URN:NBN:fi:aalto-201406252184
OEVS:
Electronic archive copy is available via Aalto Thesis Database.
Instructions

Reading digital theses in the closed network of the Aalto University Harald Herlin Learning Centre

In the closed network of Learning Centre you can read digital and digitized theses not available in the open network.

The Learning Centre contact details and opening hours: https://learningcentre.aalto.fi/en/harald-herlin-learning-centre/

You can read theses on the Learning Centre customer computers, which are available on all floors.

Logging on to the customer computers

  • Aalto University staff members log on to the customer computer using the Aalto username and password.
  • Other customers log on using a shared username and password.

Opening a thesis

  • On the desktop of the customer computers, you will find an icon titled:

    Aalto Thesis Database

  • Click on the icon to search for and open the thesis you are looking for from Aaltodoc database. You can find the thesis file by clicking the link on the OEV or OEVS field.

Reading the thesis

  • You can either print the thesis or read it on the customer computer screen.
  • You cannot save the thesis file on a flash drive or email it.
  • You cannot copy text or images from the file.
  • You cannot edit the file.

Printing the thesis

  • You can print the thesis for your personal study or research use.
  • Aalto University students and staff members may print black-and-white prints on the PrintingPoint devices when using the computer with personal Aalto username and password. Color printing is possible using the printer u90203-psc3, which is located near the customer service. Color printing is subject to a charge to Aalto University students and staff members.
  • Other customers can use the printer u90203-psc3. All printing is subject to a charge to non-University members.
Location:P1 Ark Aalto  1715   | Archive
Keywords:GPU
CUDA
OpenCL
parallel processing
energy efficient computing
high performance embedded computing
GPU compute
GPGPU
rinnakkaislaskenta
energiatehokas suorittaminen
korkean suorituskyvyn sulautettu laskenta
grafiikkasuoritinlaskenta
Abstract (eng):As the increase in the sequential processing performance of general-purpose central processing units has slowed down dramatically, computer systems have been moving towards increasingly parallel and heterogeneous architectures.
Modern graphics processing units have emerged as one of the first affordable platforms for data-parallel processing.
Due to their closed nature, it has been difficult for software developers to observe the performance and energy efficiency characteristics of the execution of applications of graphics processing units.
In this thesis, we have explored different tools and methods for observing the execution of accelerated processing on graphics processing units.
We have found that hardware vendors provide interfaces for observing the timing of events that occur on the host platform and aggregated performance metrics of execution on the graphics processing units to some extent.
However, more fine-grained details of execution are currently available only by using graphics processing unit simulators.
As a proof-of-concept, we have studied a functional graphics processing unit simulator as a tool for understanding the energy efficiency of accelerated processing.
The presented energy estimation model and simulation method has been validated against a face detection application.
The difference between the estimated and measured dynamic energy consumption in this case was found to be 5.4%.
Functional simulators appear to be accurate enough to be used for observing the energy efficiency of graphics processing unit accelerated processing in certain use-cases.
Abstract (fin):Suorittimien sarjallisen suorituskyvyn kasvun hidastuessa tietokonejärjestelmät ovat siirtymässä kohti rinnakkaislaskentaa ja heterogeenisia arkkitehtuureja.
Modernit grafiikkasuorittimet ovat yleistyneet ensimmäisinä huokeina alustoina yleisluonteisen kiihdytetyn datarinnakkaisen laskennan suorittamiseen.
Grafiikkasuorittimet ovat usein suljettuja alustoja, minkä takia ohjelmistokehittäjien on vaikea havainnoida tarkempia yksityiskohtia suorituksesta liittyen laskennan suorituskykyyn ja energian kulutukseen.
Tässä työssä on tutkittu erilaisia työkaluja ja tapoja tarkkailla ohjelmien kiihdytettyä suoritusta grafiikkasuorittimilla.
Laitevalmistajat tarjoavat joitakin rajapintoja tapahtumien ajoituksen havainnointiin sekä isäntäalustalla että grafiikkasuorittimella.
Laskennan tarkempaan havainnointiin on kuitenkin usein käytettävä grafiikkasuoritinsimulaattoreita.
Työn kokeellisessa osuudessa työssä on tutkittu funktionaalisten grafiikkasuoritinsimulaattoreiden käyttöä työkaluna grafiikkasuorittimella kiihdytetyn laskennan energiantehokkuuden arvioinnissa.
Työssä on malli grafiikkasuorittimen energian kulutuksen arviontiin.
Arvion validointiin on käytetty kasvontunnistussovellusta.
Mittauksissa arvioidun ja mitatun energian kulutuksen eroksi mitattiin 5.4%.
Funktionaaliset simulaattorit ovat mittaustemme perusteella tietyissä käyttötarkoituksissa tarpeeksi tarkkoja grafiikkasuorittimella kiihdytetyn laskennan energiatehokkuuden arviointiin.
ED:2014-08-03
INSSI record number: 49396
+ add basket
« previous | next »
INSSI