The original paper is in English. Non-English content has been machine-translated and may contain typographical errors or mistranslations. ex. Some numerals are expressed as "XNUMX".
Copyrights notice
The original paper is in English. Non-English content has been machine-translated and may contain typographical errors or mistranslations. Copyrights notice
Kami mencadangkan saluran paip penggubahan novel dan teknik pengimbangan beban dinamik untuk pemaparan volum yang menggunakan struktur kumpulan dua lapisan untuk mencapai pengimbangan beban yang berkesan dan berskala. Teknik ini membolehkan setiap proses untuk memaparkan data daripada kawasan bukan bersebelahan volum dengan kesan minimum pada jumlah masa pemaparan. Kami menunjukkan keberkesanan teknik yang dicadangkan dengan melakukan satu set eksperimen pada kelompok GPU moden. Eksperimen menunjukkan bahawa menggunakan teknik ini menghasilkan penggunaan memori terburuk sehingga 35.7% lebih rendah berbanding dengan penggunaan memori dinamik. k-d teknik pengimbangan beban pokok, sambil pada masa yang sama mencapai prestasi pemaparan yang serupa atau lebih tinggi. Teknik yang dicadangkan juga dapat mengurangkan jumlah data yang dipindahkan semasa peringkat pengimbangan beban sehingga 72.2%. Teknik ini berpotensi untuk digunakan dalam banyak senario di mana teknik pengimbangan beban dinamik lain telah terbukti tidak mencukupi, seperti semasa visualisasi berskala besar.
Marcus WALLDEN
Osaka University
Stefano MARKIDIS
KTH Royal University of Technology
Masao OKITA
Osaka University
Fumihiko INO
Osaka University
The copyright of the original papers published on this site belongs to IEICE. Unauthorized use of the original or translated papers is prohibited. See IEICE Provisions on Copyright for details.
Salinan
Marcus WALLDEN, Stefano MARKIDIS, Masao OKITA, Fumihiko INO, "Memory Efficient Load Balancing for Distributed Large-Scale Volume Rendering Using a Two-Layered Group Structure" in IEICE TRANSACTIONS on Information,
vol. E102-D, no. 12, pp. 2306-2316, December 2019, doi: 10.1587/transinf.2019PAP0003.
Abstract: We propose a novel compositing pipeline and a dynamic load balancing technique for volume rendering which utilizes a two-layered group structure to achieve effective and scalable load balancing. The technique enables each process to render data from non-contiguous regions of the volume with minimal impact on the total render time. We demonstrate the effectiveness of the proposed technique by performing a set of experiments on a modern GPU cluster. The experiments show that using the technique results in up to a 35.7% lower worst-case memory usage as compared to a dynamic k-d tree load balancing technique, whilst simultaneously achieving similar or higher render performance. The proposed technique was also able to lower the amount of transferred data during the load balancing stage by up to 72.2%. The technique has the potential to be used in many scenarios where other dynamic load balancing techniques have proved to be inadequate, such as during large-scale visualization.
URL: https://global.ieice.org/en_transactions/information/10.1587/transinf.2019PAP0003/_p
Salinan
@ARTICLE{e102-d_12_2306,
author={Marcus WALLDEN, Stefano MARKIDIS, Masao OKITA, Fumihiko INO, },
journal={IEICE TRANSACTIONS on Information},
title={Memory Efficient Load Balancing for Distributed Large-Scale Volume Rendering Using a Two-Layered Group Structure},
year={2019},
volume={E102-D},
number={12},
pages={2306-2316},
abstract={We propose a novel compositing pipeline and a dynamic load balancing technique for volume rendering which utilizes a two-layered group structure to achieve effective and scalable load balancing. The technique enables each process to render data from non-contiguous regions of the volume with minimal impact on the total render time. We demonstrate the effectiveness of the proposed technique by performing a set of experiments on a modern GPU cluster. The experiments show that using the technique results in up to a 35.7% lower worst-case memory usage as compared to a dynamic k-d tree load balancing technique, whilst simultaneously achieving similar or higher render performance. The proposed technique was also able to lower the amount of transferred data during the load balancing stage by up to 72.2%. The technique has the potential to be used in many scenarios where other dynamic load balancing techniques have proved to be inadequate, such as during large-scale visualization.},
keywords={},
doi={10.1587/transinf.2019PAP0003},
ISSN={1745-1361},
month={December},}
Salinan
TY - JOUR
TI - Memory Efficient Load Balancing for Distributed Large-Scale Volume Rendering Using a Two-Layered Group Structure
T2 - IEICE TRANSACTIONS on Information
SP - 2306
EP - 2316
AU - Marcus WALLDEN
AU - Stefano MARKIDIS
AU - Masao OKITA
AU - Fumihiko INO
PY - 2019
DO - 10.1587/transinf.2019PAP0003
JO - IEICE TRANSACTIONS on Information
SN - 1745-1361
VL - E102-D
IS - 12
JA - IEICE TRANSACTIONS on Information
Y1 - December 2019
AB - We propose a novel compositing pipeline and a dynamic load balancing technique for volume rendering which utilizes a two-layered group structure to achieve effective and scalable load balancing. The technique enables each process to render data from non-contiguous regions of the volume with minimal impact on the total render time. We demonstrate the effectiveness of the proposed technique by performing a set of experiments on a modern GPU cluster. The experiments show that using the technique results in up to a 35.7% lower worst-case memory usage as compared to a dynamic k-d tree load balancing technique, whilst simultaneously achieving similar or higher render performance. The proposed technique was also able to lower the amount of transferred data during the load balancing stage by up to 72.2%. The technique has the potential to be used in many scenarios where other dynamic load balancing techniques have proved to be inadequate, such as during large-scale visualization.
ER -