Power Management Circuitry for Data Processing Apparatus
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The increasing performance gap between highest and lowest loading in data processing apparatuses leads to the need for more powerful power supplies, increasing cost and size, while voltage and frequency scaling impacts throughput, necessitating a method to reduce power consumption without significant voltage and frequency reduction.
Innovation Solution
Implementing power management circuitry that controls voltage supply and clock signal frequency to restrict the execution of high power events, thereby managing power consumption by throttling their execution, allowing lower-rated power supplies to be used while maintaining throughput.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Use of energy by moving object
If voltage and frequency scaling is used to limit power consumption, then power consumption is reduced, but throughput is impacted as instructions are slowed down
Solution Approach 1:
The patent segments the event stream into high power events and normal events, applying different handling strategies to each. High power events are identified and tracked separately, allowing selective restriction of their execution while permitting normal events to proceed at full speed, thus reducing overall power consumption without significantly impacting throughput.
Solution Approach 2:
The patent dynamically adjusts the restriction of high power event execution based on tracked metrics. The system monitors the frequency and pattern of high power events and adaptively controls their execution rate, creating a dynamic balance between power consumption and throughput rather than applying static voltage/frequency scaling.
2Power
If more powerful power supplies are provided to handle worst-case loading, then sufficient power is available for highest loading scenarios, but cost and size of the data processing apparatus increase
Solution Approach 1:
The patent applies partial action by restricting only the execution of high power events rather than uniformly scaling back all operations. This selective approach allows the system to operate with a smaller power supply capacity than would be required for worst-case scenarios, since the high power events are throttled to prevent exceeding the reduced power supply's capacity.
3Power
If more powerful power supplies are provided to handle worst-case loading, then sufficient power is available for highest loading scenarios, but cost of the data processing apparatus increases
Solution Approach 1:
The patent applies partial action by restricting only the execution of high power events rather than uniformly scaling back all operations. This selective approach allows the system to operate with a smaller power supply capacity than would be required for worst-case scenarios, since the high power events are throttled to prevent exceeding the reduced power supply's capacity.
Data Source
AI summary
A data processing apparatus includes processing circuitry to process an event stream that includes one or more high power events. Tracking circuitry tracks the one or more high power events and power management circuitry manages power consumption by controlling a voltage supply and a frequency of a clock signal provided to the processing circuitry. The power management circuitry controls an extent to which execution by the processing circuitry of the high power events is restricted.