建構虛擬化InfiniBand和10G乙太網路於虛擬GPU叢集

Tunghai University Institutional Repository > 工學院 > 資訊工程學系所 > 碩士論文 > Item 310901/25486

Please use this identifier to cite or link to this item: http://140.128.103.80:8080/handle/310901/25486

Title:	建構虛擬化InfiniBand和10G乙太網路於虛擬GPU叢集
Other Titles:	On Construction of a Virtual GPU Cluster with InfiniBand and 10G Ethernet Virtualization
Authors:	羅育尚 Yu-Shang Lo
Contributors:	劉榮春;楊朝棟 Jung-Chun Liu;Chao-Tung Yang 資訊工程學系
Keywords:	虛擬叢集;虛擬化;CUDA;InfiniBand;VMware;HPL Virtualization;CUDA;InfiniBand;VMware;Virtual Cluster;HPL
Date:	2014
Issue Date:	2015-04-15T02:54:21Z (UTC)
Abstract:	為了因應日益龐大的計算需求，圖形處理器和CUDA，被用來建構更高效能的計算環境。由於圖型處理器的高計算效能，使它成為高效能的計算環境中不可或缺的配備。CUDA是NVIDIA公司開發的一種整合技術，適合用來撰寫高度平行化應用程式，它需要一些平行化建構的概念來使程式達到更高的效率，例如：一個有層次的線程區塊、共享式的記憶體和屏障同步…等。圖形處理器和CUDA現在也被使用在雲端的環境，為了使雲端能有更高效能的計算能力。在雲端架構中，虛擬化是很重要的一環。藉由讓虛擬機器使用NVIDIA的顯示卡透過CUDA去提升計算效能，這使虛擬機可以同時使用虛擬的中央處理器和實體的圖形處理器來做運算。InfiniBand是一種比乙太網路還快的傳輸媒介，以前虛擬機只能透過間接的方式去使用InfiniBand，現在有許多虛擬化平台可以讓虛擬機直接使用InfiniBand，這必然會使虛擬機間的傳輸速度改善許多。本論文中使用多張的圖形處理器在虛擬化平台上建構一個高效能的叢集運算環境，並透過HPL來比較直接使用InfiniBand和間接使用InfiniBand的傳輸效能。本論文中我們打算選用大家較熟悉的虛擬化平台VMware來做測試，透過使用GPU passthrough和InfiniBand和10G網卡的虛擬化或是passthrough來達到加快虛擬化叢集環境的目的。 Due to increasing requirement of computing capability, the graphics processor unit and CUDA are used to build a higher-performance computing environment. The graphics processor unit is necessary for building the high-performance computing environment because of its high computing performance. CUDA, a parallel computing platform and programming model created by NVIDIA, utilizes some parallel construct concepts to upgrade performance, such as hierarchical thread blocks, shared memory, and barrier synchronization. The graphics processor unit and CUDA are also used in cloud computing, because they can provide high-performance computing capabilities. Virtualization plays a very important part in the cloud architecture. Virtual machines built with the NVIDIA graphics card can use CUDA to provide virtual machine computing capability. That makes virtual machine have not only virtual CPUs, but also physical graphics processors to do computations. InfiniBand is faster than Ethernet as the transmission medium. In the past, virtual machine cannot use direct InfiniBand, but now, many virtualization platforms can do it, that brings transmission speed improvement between virtual machines. In this work, we use many graphics processors units to build a high-performance computing cloud cluster. Then, we compare performance of using direct InfiniBand with that of using indirect InfiniBand transmission performance by running HPL. In this thesis, we use a well-known virtualization platform, i.e., VMware to do experiments in this paper. And then, we use GPU passthrough and InfiniBand virtual and 10G Ethernet passthrough to improve performance of the virtual cluster.
Appears in Collections:	[資訊工程學系所] 碩士論文

Files in This Item:

File	Size	Format
102THU00394015-001.pdf	1263Kb	Adobe PDF	470	View/Open

Loading...