Idle State Configuration via Idle Duration Prediction

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Transitioning computing device entities into and out of idle states incurs delays and inefficiencies, particularly when idle periods are short, leading to unnecessary power consumption and operational inefficiencies due to the costs of preserving and restoring architectural state.

Innovation Solution

A computing device uses an idle duration history and history table to predict the duration of upcoming idle periods, determining the appropriate idle state for entities based on these predictions, thereby avoiding unnecessary transitions and power expenditure by selecting idle states that do not require preserving architectural state when periods are short.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Use of energy by moving object

If an entity transitions into a deep idle state (e.g., C6) to conserve power, then power consumption is reduced, but transition delays increase and operational efficiency deteriorates

Engineering Contradiction:

Improvepower consumptionVSAvoidtransition delay

Core Design Contradiction:

Use of energy by moving objectVSLoss of time

Solution Approach 1:

The system dynamically selects among multiple idle states (C0-C6) based on predicted idle period duration. Instead of always transitioning to the deepest idle state, the controller adapts the idle state selection to match the expected duration, optimizing the trade-off between power savings and transition overhead. This dynamic adaptation resolves the contradiction by making the system flexible rather than fixed in its idle state choice.

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

The system changes the operational parameters (voltage levels, clock frequencies) of idle states based on the predicted idle period length. For short idle periods, shallower idle states with higher voltage/clock parameters are selected to minimize transition delays. For longer idle periods, deeper idle states with lower parameters are chosen to maximize power savings. This parameter adjustment resolves the contradiction by matching system configuration to operational context.

Inventive Principle:

Principle #35Parameter changes

2Use of energy by moving object

If an entity transitions into an idle state that preserves architectural state, then power consumption is reduced during long idle periods, but transition delays increase due to state preservation operations

Engineering Contradiction:

Improvepower consumptionVSAvoidoperational efficiency

Core Design Contradiction:

Use of energy by moving objectVSProductivity

Solution Approach 1:

The system dynamically adjusts whether to preserve architectural state based on the predicted idle period duration. For short idle periods, the system skips state preservation operations to maintain operational efficiency. For long idle periods, state preservation is enabled to maximize power savings. This dynamic decision-making resolves the contradiction by adapting the state preservation behavior to the specific operational context.

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

The system changes the configuration parameter of architectural state preservation (enabled/disabled) based on idle period predictions. When transitioning to idle states for short durations, the preservation parameter is set to disabled to avoid delays. For longer durations, it is enabled to achieve greater power savings. This parameter change resolves the contradiction by optimizing the trade-off between power consumption and operational efficiency.

Inventive Principle:

Principle #35Parameter changes

3Adaptability or versatility

If an entity frequently transitions between idle states, then power management flexibility is improved, but unnecessary power expenditure occurs due to repeated transitions

Engineering Contradiction:

Improvepower management flexibilityVSAvoidpower expenditure

Core Design Contradiction:

Adaptability or versatilityVSLoss of energy

Solution Approach 1:

The system performs preliminary action by predicting the duration of upcoming idle periods before making transition decisions. This advance prediction allows the system to plan optimal idle state selections and avoid unnecessary transitions. By knowing in advance whether an idle period will be short or long, the system can prepare the appropriate idle state configuration, resolving the contradiction between flexibility and energy waste.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system uses feedback from actual idle period durations to refine future predictions and adjust idle state selection strategies. By continuously learning from past idle period patterns, the system improves its ability to predict future idle durations accurately, enabling better transition decisions that balance flexibility with energy conservation. This feedback mechanism resolves the contradiction by making the system progressively more efficient.

Inventive Principle:

Principle #23Feedback

Data Source

PatentUS9471130B2Configuring idle states for entities in a computing device based on predictions of durations of idle periods

Publication Date: 2016.10.18 ADVANCED MICRO DEVICES INC

AI summary

The described embodiments include a computing device with an entity (a processor, a processor core, etc.) and a controller. In these embodiments, the controller, using an idle duration history, predicts a duration of a next idle period for the entity. Based on the predicted duration of the next idle period, the controller configures the entity to operate in a corresponding idle state.