The Experts below are selected from a list of 1176 Experts worldwide ranked by ideXlab platform
Yue Zha - One of the best experts on this subject based on the ideXlab platform.
-
hetero vital a Virtualization Stack for heterogeneous fpga clusters
International Symposium on Computer Architecture, 2021Co-Authors: Yue ZhaAbstract:With field-programmable gate arrays (FPGAs) being widely deployed into data centers, an efficient Virtualization support is required to fully unleash the potential of cloud FPGAs. Nevertheless, existing FPGA Virtualization solutions only support a homogeneous FPGA cluster comprising identical FPGA devices. Representative work such as ViTAL provides sufficient system support for scale-out acceleration and improves the overall resource utilization through a fine-grained spatial sharing. While these existing solutions (including ViTAL) can efficiently virtualize a homogeneous cluster, it is hard to extend them to virtualizing a heterogeneous cluster which comprises multiple types of FPGAs. We expect the future cloud FPGAs are likely to be more heterogeneous due to hardware rolling upgrade. In this paper, we rethink FPGA Virtualization from ground up and propose HETERO-VITAL to virtualize heterogeneous FPGA clusters. We identify the conflicting requirements of runtime management and offline compilation when designing the abstraction for a heterogeneous cluster, which is also the fundamental reason why the single-level abstraction as proposed in ViTAL (and other prior works) cannot be trivially extended to the heterogeneous case. To decouple these conflicting requirements, we provide a two-level system abstraction in HETERO-VITAL. Specifically, the high-level abstraction is FPGA-agnostic and provides a simple and homogeneous view of the FPGA resources to simplify the runtime management. On the contrary, the low-level abstraction is FPGA-specific and exposes sufficient spatial resource constraints to the compilation framework to ensure the mapping quality. Rather than simply adding a layer on top of the single-level abstraction as proposed in ViTAL and other prior work, we judiciously determine how much hardware details should be exposed at each level to balance the management complexity, mapping quality and compilation cost. We then develop a compilation framework to map applications onto this two-level abstraction with several optimization techniques to further improve the mapping quality. We also provide a runtime management policy to alleviate the fragmentation issue, which becomes more severe in a heterogeneous cluster due to the distinct resource capacities of diverse FPGAs. We evaluate HETERO-VITAL on a custom-built FPGA cluster and demonstrate its effectiveness using machine learning and image processing applications. Results show that HETERO-VITAL reduces the average response time (a critical metric for QoS) by 79.2% for a heterogeneous cluster compared to the non-virtualized baseline. When virtualizing a homogeneous cluster, HETERO-VITAL also reduces the average response time by 42.0% compared with ViTAL due to a better system design.
Shabaitah, Anas Raqi - One of the best experts on this subject based on the ideXlab platform.
-
Server-Based Desktop Virtualization
RIT Scholar Works, 2014Co-Authors: Shabaitah, Anas RaqiAbstract:Virtualization can be accomplished at different layers in the computational Stack and with different goals (servers, desktops, applications, storage and network). This research focuses on server-based desktop Virtualization. According to the Gartner group, the main business drivers for adopting desktop Virtualization are: application compatibility, business continuity, security and compliance, mobility and improved productivity [15]. Despite these business drivers, desktop Virtualization has not been widely adopted. According to a survey conducted by Matrix42, only 5% of desktop computers are virtualized [37]. The research deals with the challenges preventing the wider adoption of server-based desktop Virtualization while focusing on two of the main Virtualization architectures: session-based desktop Virtualization (SBDV) and virtual desktop infrastructure (VDI). The first chapter introduces some of the challenges faced by large organizations in their efforts to create a cost effective and manageable desktop computing environment. The second chapter discusses two of the main server-based desktop Virtualizations (VDI and SBDV), illustrating some of the advantages and disadvantages in these different architectures. The third chapter focuses on some of the technical challenges and provides recommendations regarding server-based desktop Virtualization. In the fourth chapter, measurements are conducted for the utilization and performance of SBDV on different 3 user profiles (light, heavy and multimedia). Data and results collected from desktop assessment and lab are used to formulate baselines and metrics for capacity planning. According to the conducted measurements, it is concluded that light and heavy profiles can be virtualized using SBDV, while for multimedia profiles, additional capacity planning and resource allocation are required. Multimedia profiles can be virtualized with VDI considering client-side rendering to avoid network bandwidth congestion. While the research focuses on VDI and SBDV, it highlights few points related to client access devices (CADs). CADs are one of the main components in the desktop Virtualization Stack (OS Virtualization, session Virtualization, application Virtualization, connection broker, CADs and user data and profiles). The latter chapter of the research focuses on conclusions and future work toward greater levels of adoption of VDI and SBDV
Muhammed Naazer Ashraf - One of the best experts on this subject based on the ideXlab platform.
-
scratching the surface of windows server 2016 and system center configuration manager current branch
SIGUCCS: User Services Conference, 2017Co-Authors: Muhammed Naazer AshrafAbstract:Lehigh University has set a goal to implement System Center Configuration Manager by the end of 2017. This project is being spearheaded by one of our Senior Computing Consultants who has been researching and trained in the Microsoft Virtualization Stack. We will discuss our roadmaps, results from our proof-of-concept environments, and discussions in driving this project.
