Dynamic Guard Band Adjusting Droop Sensor Sensitivity

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

In distributed computing environments, processor cores face challenges in maintaining optimal timing and performance due to voltage droops, leading to unrecoverable errors and performance degradation, which existing methods address inadequately by relying on static guard bands that increase power consumption and reduce yield.

Innovation Solution

A dynamic guard band system that adjusts droop sensor calibration and tuning parameters based on core recovery events, reducing timing and voltage margins during light workloads and increasing them during heavy workloads to prevent errors and optimize performance.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If a static guard band is used to maintain timing margin, then processor reliability is improved, but power consumption increases and yield decreases

Engineering Contradiction:

Improveprocessor reliabilityVSAvoidpower consumption

Core Design Contradiction:

ReliabilityVSUse of energy by moving object

Solution Approach 1:

The patent implements a dynamic guard band mechanism that adjusts timing margins based on real-time workload conditions. The system monitors workload intensity and dynamically modifies the guard band value, transitioning from a fixed static guard band to a variable dynamic guard band that adapts to current operational demands, thereby reducing power consumption during light workloads while maintaining reliability during heavy workloads

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

The system changes the timing margin parameter dynamically based on workload conditions. By monitoring workload intensity and adjusting the guard band parameter accordingly, the system optimizes the balance between reliability and power consumption, using larger timing margins only when necessary for heavy workloads and reducing them during lighter operations

Inventive Principle:

Principle #35Parameter changes

2Reliability

If a static guard band is used to maintain timing margin, then processor reliability is improved, but the number of processor cores running concurrently decreases

Engineering Contradiction:

Improveprocessor reliabilityVSAvoidprocessor core concurrency

Core Design Contradiction:

ReliabilityVSProductivity

Solution Approach 1:

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

Inventive Principle:

Principle #35Parameter changes

3Use of energy by moving object

If timing margin is reduced to save power, then power consumption decreases, but the processor becomes more susceptible to unrecoverable errors

Engineering Contradiction:

Improvepower consumptionVSAvoiderror susceptibility

Core Design Contradiction:

Use of energy by moving objectVSReliability

Solution Approach 1:

The patent implements a feedback mechanism that continuously monitors workload conditions and adjusts the guard band accordingly. The system uses workload monitoring feedback to determine when to reduce or maintain timing margins, ensuring that power consumption is optimized without compromising reliability, as the guard band is only reduced during periods of low workload stress

Inventive Principle:

Principle #23Feedback

Solution Approach 2:

Inventive Principle:

Principle #15Dynamics

4Reliability

If droop sensor sensitivity is increased to detect voltage droops earlier, then timing protection is improved, but the rate of recovery events increases impacting performance

Engineering Contradiction:

Improvetiming protectionVSAvoidperformance

Core Design Contradiction:

ReliabilityVSProductivity

Solution Approach 1:

The patent dynamically adjusts the droop sensor calibration parameter based on workload conditions. By modifying the sensor sensitivity parameter in response to monitored workload intensity, the system optimizes the balance between early voltage droop detection and performance, reducing false recovery events during light workloads while maintaining protective detection during heavy workloads

Inventive Principle:

Principle #35Parameter changes

Data Source

PatentUS11953982B2Dynamic guard band with timing protection and with performance protection

Publication Date: 2024.04.09 INTERNATIONAL BUSINESS MACHINE CORPORATION

AI summary

Embodiments include in response to monitoring a processor during operation, detecting a first number of core recovery events in the processor, determining that the first number of core recovery events fulfills a first condition for the first core recovery events threshold, and modifying a value of at least one droop sensor parameter of the processor by a first amount. The at least one droop sensor parameters affects a sensitivity to a voltage droop. In response to modifying the value of the droop sensor parameter by the first amount, a second number of core recovery events is detected in the processor. It is determined that the second number of core recovery events fulfills a second condition for a second core recovery events threshold, and the value of the at least one droop sensor parameter is modified by a second amount.