Dynamic Accelerator Image Loading for NFV Resource Efficiency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing NFV architectures face performance degradation when handling multiple network functions, leading to resource wastage due to the inability of existing accelerators to dynamically load different acceleration functions, requiring multiple hardware accelerators for each function.
Innovation Solution
An accelerator loading method and system that dynamically loads images meeting specific acceleration requirements for virtual machines, utilizing an accelerator image repository and acceleration database to allocate and manage available accelerators efficiently, allowing a single accelerator to implement different functions by loading various images.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If multiple hardware accelerators are provided for different acceleration functions, then the acceleration function requirements are met, but hardware resources are wasted
Solution Approach 1:
The patent implements a universal accelerator architecture that can load different images to provide multiple acceleration functions. Instead of having separate dedicated hardware accelerators for each function, a single accelerator device can dynamically load appropriate images to deliver the required acceleration function, thereby reducing hardware resource waste while maintaining versatility.
Solution Approach 2:
The patent introduces dynamic image loading capability to the accelerator, allowing it to switch between different acceleration functions by loading different images at runtime. This dynamic reconfiguration enables the accelerator to adapt to different acceleration requirements without requiring multiple static hardware devices, resolving the contradiction between versatility and resource efficiency.
2Reliability
If a single accelerator loads a specific function image, then the acceleration function is implemented, but the accelerator cannot dynamically load different images for different functions
Solution Approach 1:
The patent transforms the static accelerator into a dynamic system by introducing an image loading mechanism. The accelerator can now load different images corresponding to different acceleration functions based on runtime requirements, while maintaining reliable function implementation through proper image validation and loading protocols.
Solution Approach 2:
The patent introduces an image as an intermediary between the accelerator hardware and the acceleration function. The image serves as a configurable layer that enables the accelerator to implement different functions without changing the underlying hardware, thus achieving both reliability and adaptability.
3Adaptability or versatility
If NFV architecture uses standard hardware, then hardware unification is achieved, but performance deteriorates when processing large quantities of network functions
Solution Approach 1:
The patent applies local quality enhancement by introducing hardware acceleration capabilities to specific network functions within the NFV architecture. Instead of requiring all hardware to be high-performance, only the critical acceleration functions receive dedicated hardware support through loaded images, while other functions continue to use standard hardware processing.
Solution Approach 2:
The patent creates a composite processing architecture that combines standard hardware with hardware-accelerated functions. The system uses a mixture of software-based processing for general functions and hardware-based acceleration for performance-critical functions, achieving both unification and high performance.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The present invention provides an accelerator loading method and system. An accelerator loading apparatus obtains an acceleration requirement, where the acceleration requirement includes an acceleration function of a to-be-created virtual machine and acceleration performance of the to-be-created virtual machine. The accelerator loading apparatus determines an image that meets the acceleration function and the acceleration performance, and determines a target host in which an available accelerator that can load the image is located. The accelerator loading apparatus sends an image loading command to the target host. The image loading command includes a descriptor of the image, and the image loading command is used to enable the target host to load the image for the available accelerator based on the image loading command. In the present invention, a target host that can create the virtual machine may be determined based on the acceleration function and the acceleration performance of the to-be-created virtual machine, and an image used for acceleration is loaded to an available accelerator of the target host, to implement dynamic accelerator loading and deployment.