Sökning: "mpi"

Visar resultat 21 - 25 av 64 uppsatser innehållade ordet mpi.

  1. 21. Stämmer ischemidrabbat koronarartärsområde visat med myokardscintigrafi överens med koronarartärsmorfologi visat med koronarangiografi

    Kandidat-uppsats, Högskolan i Jönköping/HHJ, Avd. för naturvetenskap och biomedicin

    Författare :Carl Bartoli; Magnus Nordlöf; [2018]
    Nyckelord :Koronar revaskularisering; polar plot; perfusion.;

    Sammanfattning : Myocardial Perfusion Imaging (MPI) and Coronary Angiography (CA) are two methods that can be used in the investigation of ischemic heart disease. The purpose of this study is to compare medical answers from MPI and CA to find out if suspected coronary artery disease shown with MPI correlates with coronary artery morphology shown with CA, and which treatment is given based on Summed Difference Percent (SD%). LÄS MER

  3. 22. Effects of Thread-Local Task Trees in the CHT-MPI C++ Programming Library for the Chunks and Tasks Programming Model

    Kandidat-uppsats, Uppsala universitet/Institutionen för informationsteknologi

    Författare :Joseph Rogers; [2018]
    Nyckelord :;

    Sammanfattning : Programming efficiently in extremely high thread count environments can be both challenging and tedious for the modern programmer. Task based programming models provide an abstraction of the low--level system details by instead allowing a program to be expressed as a set of tasks to executed by the system. LÄS MER

  4. 23. Spark for HPC: a comparison with MPI on compute-intensive applications using Monte Carlo method

    Master-uppsats, Uppsala universitet/Institutionen för informationsteknologi

    Författare :Shen Huije; Tingwei Huang; [2018]
    Nyckelord :;

    Sammanfattning : With the emergence of various big data platforms in recent years, Apache Spark - a distributed large-scale computing platform, is perceived as a potential substitute for Message Passing Interface (MPI) in High Performance Computing (HPC). Due to the limitations in fault-tolerance, dynamic resource handling and ease of use, MPI, as a dominant method to achieve parallel computing in HPC, is often associated with higher development time and costs in enterprises such as Scania IT. LÄS MER

  5. 24. Analysis and Comparison of Distributed Training Techniques for Deep Neural Networks in a Dynamic Environment

    Uppsats för yrkesexamina på avancerad nivå, KTH/Skolan för elektroteknik och datavetenskap (EECS)

    Författare :Ermias Gebremeskel; [2018]
    Nyckelord :deep learning; large scale distributed deep learning; data parallelism;

    Sammanfattning : Deep learning models' prediction accuracy tends to improve with the size of the model. The implications being that the amount of computational power needed to train models is continuously increasing. Distributed deep learning training tries to address this issue by spreading the computational load onto several devices. LÄS MER

  7. 25. Effective Data Redistribution and Load Balancing for Sort-Last Volume Rendering Using a Group Hierarchy

    Uppsats för yrkesexamina på avancerad nivå, KTH/Skolan för elektroteknik och datavetenskap (EECS)

    Författare :Marcus Walldén; [2018]
    Nyckelord :volume rendering; cuda; load balancing; sort-last; mpi;

    Sammanfattning : Volumetric rendering is used to visualize volume data from e.g. scientific simulations. Many advanced applications use large gigabyte- or terabyte-sized data sets, which typically means that multiple compute nodes need to partake in the rendering process to achieve interactive frame rates. LÄS MER