vGPU Scheduler for Network Function Placement

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

The high cost of network function virtualization makes it more expensive than traditional dedicated hardware solutions, necessitating innovation to improve efficiency and reduce costs in vGPU-enabled environments.

Innovation Solution

A scheduler-based mechanism for network function placement in vGPU-enabled environments that optimizes the allocation of network functions across virtual machines and GPUs, using vGPU scheduling policies, network function placement rules, and machine learning to ensure efficient resource utilization and minimize data transfer, while considering trust status and resource availability.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If network functions are virtualized using distributed infrastructure, then hardware deployment flexibility is improved, but cost increases beyond dedicated hardware solutions

Engineering Contradiction:

Improvehardware deployment flexibilityVSAvoidcost

Core Design Contradiction:

Adaptability or versatilityVSQuantity of substance

Solution Approach 1:

The patent merges multiple network function virtualization operations onto a single GPU device, allowing multiple virtual machines to share GPU resources for network function processing. This consolidation reduces the number of separate hardware devices needed and lowers overall costs while maintaining deployment flexibility.

Inventive Principle:

Principle #5Merging (Combining)

Solution Approach 2:

The GPU device is designed to perform multiple network function virtualization tasks simultaneously, serving as a universal platform for different network functions across multiple virtual machines. This multi-functionality eliminates the need for dedicated hardware for each network function, reducing costs while preserving adaptability.

Inventive Principle:

Principle #6Universality (Multi-functionality)

2Productivity

If network functions are placed on vGPU-enabled devices, then resource utilization efficiency is improved, but data transfer overhead increases

Engineering Contradiction:

Improveresource utilization efficiencyVSAvoiddata transfer overhead

Core Design Contradiction:

ProductivityVSLoss of energy

Solution Approach 1:

The patent segments the GPU device into multiple virtual GPU instances, each assigned to specific virtual machines. This segmentation allows network functions to be placed close to the data sources they process, reducing cross-VM data transfer overhead while maintaining high resource utilization through shared GPU resources.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent introduces a network function virtualization layer that acts as an intermediary between virtual machines and the physical GPU device. This intermediary optimizes data transfer paths and manages resource allocation, reducing overhead while improving overall system efficiency.

Inventive Principle:

Principle #24Intermediary (Mediator)

Data Source

PatentUS12124867B2Network function placement in vGPU-enabled environments

Publication Date: 2024.10.22 VMWARE INC

AI summary

Disclosed are aspects of network function placement in virtual graphics processing unit (vGPU)-enabled environments. In one example a network function request is associated with a network function. A scheduler selects a vGPU-enabled GPU to handle the network function request. The vGPU-enabled GPU is selected in consideration of a network function memory requirement or a network function IO requirement. The network function request is processed using an instance of the network function within a virtual machine that is executed using the selected vGPU-enabled GPU.