Adaptive Energy Optimal Computing Control System
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Integrated circuits, such as SoCs, face variations in performance due to temperature and semiconductor process variations, leading to inefficiencies in power consumption, particularly in IoT applications where low power consumption and high efficiency are critical for battery life. Existing techniques fail to effectively manage power consumption across varying temperatures and usage scenarios, and do not adequately account for leakage power at low voltage levels.
Innovation Solution
An energy-optimal computing control system dynamically adjusts clock rates and supply voltage to minimize total energy consumption, operating near the optimal voltage close to the threshold voltage to maximize energy efficiency. This involves a simplified model for estimating chip-level leakage current and using a fractional divider for fine-resolution frequency control, along with a look-up table or dynamic calculation to adjust supply voltage based on temperature.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Use of energy by moving object
If the supply voltage is reduced to minimize dynamic power consumption, then energy efficiency improves, but leakage power becomes dominant and circuit timing requirements cannot be met
Solution Approach 1:
The patent dynamically adjusts the supply voltage to an optimal value that balances dynamic and leakage power consumption. Instead of using fixed voltage levels, the system continuously monitors operating conditions and modifies the voltage parameter to minimize total power consumption while ensuring timing requirements are met and leakage power does not become dominant.
2Speed
If the supply voltage is increased to meet timing requirements, then circuit performance improves, but energy consumption increases
Solution Approach 1:
The patent implements dynamic voltage adjustment where the supply voltage is continuously adapted based on real-time monitoring of circuit performance and operating conditions. The voltage is increased only when necessary to meet timing requirements and decreased when performance allows, creating a dynamic balance between speed and energy consumption rather than using a static voltage level.
Solution Approach 2:
The system dynamically changes the supply voltage parameter to optimize the trade-off between circuit speed and energy consumption. By adjusting the voltage to an optimal value based on operating conditions, the system achieves the minimum voltage required for correct operation, thereby minimizing energy consumption while maintaining acceptable performance.
3Device complexity
If fixed voltage levels are used to simplify power management, then device complexity is reduced, but energy efficiency deteriorates due to inability to adapt to varying conditions
Solution Approach 1:
The patent implements a self-adjusting power management system that automatically monitors operating conditions and adjusts the supply voltage without requiring complex external control. The system uses on-chip sensors and control logic to autonomously optimize voltage levels, eliminating the need for complex external power management circuitry while maintaining high energy efficiency.
Data Source
AI summary
Methods, devices and systems are described that relate to an energy optimal computing control system, where clock rates and supply voltage are control knobs to adjust the total energy consumption of an electronic circuit. An ultra-wide voltage range, such as from near-threshold voltage to the device maximum voltage, is used to maximize the performance and to minimize the energy consumption. One example device includes a processor that receives or determines the temperature of the electronic circuit, and determines the optimum voltage levels and clock rates for the electronic circuit at the operating temperature. This information is provided to a voltage regulator and a clock generator to adjust the supply voltage and clock frequency accordingly.