Processor MEP Tracking Using Sensor-Driven Heuristic Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current techniques fail to efficiently identify and track the minimum energy point (MEP) in digital integrated circuits, leading to suboptimal energy consumption due to their reliance on recursive and computationally expensive methods for dynamic voltage and frequency adjustments.
Innovation Solution
A sensor-driven MEP control circuit that leverages heuristic techniques and pre-characterized lookup tables to quickly determine and update the optimal MEP operating point, enabling fast and efficient energy computation and adjustment in real-time.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If recursive and computationally expensive methods are used for dynamic voltage and frequency adjustments, then measurement precision of energy consumption is improved, but productivity of MEP identification deteriorates
Solution Approach 1:
The patent pre-characterizes the processor across process, voltage, and temperature (PVT) variations to build lookup tables containing MEP information before runtime. This preliminary characterization allows the system to quickly retrieve pre-computed MEP values without performing expensive recursive calculations during operation, thus maintaining measurement precision while dramatically improving productivity.
Solution Approach 2:
The patent creates simplified models and lookup tables that copy essential MEP characteristics from extensive pre-characterization data. Instead of performing full recursive energy computations at runtime, the system uses these copied representations (lookup tables with pre-computed MEP values) to rapidly determine optimal operating points, resolving the contradiction between precision and speed.
2Measurement precision
If extensive voltage and frequency sweeps are performed to identify MEP, then measurement precision is improved, but loss of time increases
Solution Approach 1:
The patent performs voltage and frequency sweeps in advance during processor characterization, storing the results in lookup tables. This preliminary action captures all necessary MEP information before the processor enters runtime operation, eliminating the need for time-consuming sweeps during actual operation while maintaining full measurement precision through the pre-collected data.
Solution Approach 2:
The patent implements dynamic MEP tracking that adapts to changing processor conditions (temperature, workload, process variations) by selectively accessing appropriate pre-characterized data from lookup tables. This dynamic approach maintains accuracy across varying conditions without requiring repeated static sweeps, thus reducing time loss while preserving measurement precision.
3Device complexity
If conventional techniques are used for MEP identification, then device complexity is reduced, but use of energy increases
Solution Approach 1:
The patent uses lookup tables that copy pre-computed energy consumption data and MEP information from extensive characterization. This allows the control circuit to determine energy-efficient operating points without performing complex real-time energy calculations, thereby maintaining low device complexity while achieving significant energy savings through informed voltage and frequency selection.
Solution Approach 2:
The patent changes the operating parameters (voltage and frequency) based on pre-characterized energy consumption data stored in lookup tables. By selecting operating points from this pre-analyzed data rather than computing energy consumption in real-time, the system achieves better energy efficiency while keeping the control circuit complexity manageable.
Data Source
AI summary
In one embodiment, a processor includes a minimum energy point (MEP) controller to: generate a change in thermal tracking information, based at least in part on prior and current thermal information; generate a change in activity tracking information, based at least in part on prior activity information and current activity information; and determine a MEP performance state based at least in part on the change in thermal tracking information and the change in activity tracking information. Other embodiments are described and claimed.