Recent years have seen rapid adoption of GPUs in various types of platforms because of the tremendous throughput powered by massive parallelism. However, as the computing power of GPU continues to grow at a rapid pace, it also becomes harder to utilize these additional resources effectively with the support of GPU sharing. In this work, we designed and implemented Gemini, a user-space runtime scheduling framework to enable fine-grained GPU allocation control with support for multi-tenancy and elastic allocation, which are critical for cloud and resource providers. Our key idea is to introduce the concept of kernel burst, which refers to a group of consecutive kernels launched together without being interrupted by synchronous events. Based on the characteristics of kernel burst, we proposed a low overhead event-driven monitor and a dynamic time-sharing scheduler to achieve our goals. Our experiment evaluations using five types of GPU applications show that Gemini enabled multi-tenant and elastic GPU allocation with less than 5% performance overhead. Furthermore, compared to static scheduling, Gemini achieved 20%~30% performance improvement without requiring prior knowledge of applications.

Network function virtualization (NFV) enables network operators to save costs and flexibility by replacing dedicated hardware with software network functions running on commodity servers. There is a high need for network acceleration to achieve performance comparable to hardware, which is vital for the implementation of NFV. The necessity of NFV acceleration stems from the lengthy packet delivery path following virtualization and the unavailability of generic operating system designs to serve network-specific scenarios. Therefore, the software approach alters the operating system’s processing architecture through Kernel Bypass or offload packet processing to hardware. A typical classification scheme divides it into two main categories based on technology with software and hardware. Only these two categories can be utilized to rapidly and easily establish a classification system. However, it is difficult to suggest the specifics and peculiarities of any acceleration approach during real-world operation. For a more comprehensive classification of NFV acceleration, we refer to the ETSI NFV architectural framework in this research. As the framework clearly illustrates, the technical infrastructure layer of NFV and the corresponding management roles provides a comprehensive and intuitive view of the differences between these acceleration technologies, solutions, and initiatives. Additionally, we conducted an analysis to identify opportunities for improvement in existing solutions and propose new research programs. We expect that NFV will increasingly rely on cloud services in the future. Since cloud services do not offer a choice of hardware, our acceleration method will be primarily software-based.

Cloud computing and High-Performance Computing (HPC) have become emerging paradigms that drive innovation and reshape the landscape of Information and Communication Technology (ICT)

Network function virtualization (NFV) is an emerging paradigm that has revolutionized the deployment and provisioning of telecommunication network services (NS). In NFV, a network service is implemented as a chain of network functions (NFs). NFs are virtualized and packaged into virtual network functions (VNF), so that they can be decoupled from the physical middleboxes and dynamically deployed on standard virtualization platforms in cloud or datacenters. Therefore, NFV has the potential to provide significant cost reductions in operating expenses (OPEX) and capital expenses (CAPEX) and increased agility for network service deployments. However, to build and manage a network function virtualization infrastructure (NFVI) is challenging. One of the key problems is to determine the mapping between underlying physical resources and VNFs in order to minimize energy consumption without violating service-level agreements (SLA). Many previous studies use embedding solutions to solve the problem by only considering the resource demand of NS and the resource capacity of physical network at a given time instance. But in practice, the resource demands of NS can vary over time, hence the embedding decision must be reconfigured accordingly. Many techniques can be used for reconfiguration, including resource scaling, instance migration and traffic re-direction. But each of these operation is associated with some costs, such as service interruptions, and additional resource or energy consumption. In this work, we investigate various reconfiguration strategies, formulate their costs by designing a model in the NFV placement problem, and propose Integer Linear Programming (ILP) and heuristic solutions, called REAP. In our evaluation, we observed 60% and 25% cost reduction for energy and reconfiguration costs, and we analyze the cost reduction from each of the reconfiguration methods to validate the design of our heuristic algorithm.