Power Management Virtual Machines for VM Entitlement Control

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Hypervisors currently prohibit virtual machines (VMs) from configuring power management technologies at VM host processors, preventing potential instability and ensuring control for virtualization host providers, but this restricts VM operators from optimizing energy consumption and performance.

Innovation Solution

Exposing power management technologies to VMs with a power management entitlement, allowing them to modify processor settings such as idle states and dynamic frequency scaling, while maintaining regulatory control through para-virtual services and architectural interfaces.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If hypervisors prohibit VMs from configuring power management technologies, then system stability and control are ensured, but VM operators cannot optimize energy consumption and performance

Engineering Contradiction:
Improvesystem stabilityVSAvoidpower management optimization capability
Core Design Contradiction:
ReliabilityVSAdaptability or versatility

Solution Approach 1:

The system segments power management control by introducing entitlement levels: some VMs are granted power management entitlement while others are not. This segmentation allows stability-critical VMs to remain restricted while enabling optimization for VMs that benefit from power management control, resolving the contradiction between universal stability and selective optimization capability

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The hypervisor acts as an intermediary that mediates between VM power management requests and actual processor control. When a VM with entitlement requests power management configuration, the hypervisor validates the request and implements it through controlled interfaces, maintaining system stability while enabling VM-level optimization

Inventive Principle:
Principle #24Intermediary (Mediator)

2Use of energy by moving object

If VMs are allowed to modify processor power management settings, then energy consumption can be reduced and performance boosted, but system stability may be compromised

Engineering Contradiction:
Improveenergy consumptionVSAvoidsystem stability
Core Design Contradiction:
Use of energy by moving objectVSReliability

Solution Approach 1:

The system dynamically adjusts power management settings based on VM workload characteristics and entitlement status. The hypervisor monitors system state and VM performance, allowing power management modifications only when conditions are favorable, thus enabling energy optimization while maintaining stability through adaptive control

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The system implements feedback mechanisms where the hypervisor monitors the effects of power management changes on system stability and performance. If instability is detected, the hypervisor can revert changes or prevent further modifications, creating a closed-loop control system that optimizes energy consumption while maintaining reliability

Inventive Principle:
Principle #23Feedback

Data Source

PatentUS20240354138A1Power-management virtual machines
Publication Date: 2024.10.24 MICROSOFT TECHNOLOGY LICENSING LLC
  • US20240354138A1 patent drawing
  • US20240354138A1 patent drawing
  • US20240354138A1 patent drawing

AI summary

Method, systems, and computer program products for managing a processor power management feature on behalf of a virtual machine (VM). A device may determine that a VM operating at a computer system possesses a power management entitlement. The device may identify an architectural power management feature available at a physical processor core of a processor system. The physical processor core is associated with a virtual processor core of the VM. Based on the VM possessing the power management entitlement, the device may present an interface to the VM. The interface exposes the architectural power management feature to the VM. The device may identify a request from the VM to modify a state of the architectural power management feature. Based on the request, the device may modify the state of the architectural power management feature at the physical processor core.