Dynamic Power Budget Allocation for Multi-Processor Systems
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
High-performance laptops face inefficiencies in power management, where the GPU operates at maximum power limits while the CPU consumes less than its limit, leading to unused power within the overall power budget, which hinders performance and user experience.
Innovation Solution
A method that detects the power usage of a first processor, determines extra power available within its budget, and transmits this information to a driver associated with a second processor, allowing the driver to adjust operating parameters, such as overclocking, to utilize the extra power for improved performance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the power budget is allocated to allow both CPU and GPU to operate at their maximum power limits simultaneously, then the system can handle worst-case power consumption scenarios, but during typical operating scenarios the CPU operates below its maximum power limit while the GPU operates at or near its maximum, leaving available power unused
Solution Approach 1:
The patent implements dynamic power budget allocation where the CPU's power budget is not fixed but can be temporarily reallocated to the GPU based on real-time power consumption monitoring. When the CPU operates below its maximum power limit, the unused power is dynamically transferred to the GPU, allowing the GPU to exceed its normal power limit and operate at higher performance levels.
Solution Approach 2:
The system changes the power consumption parameter of the GPU by allowing it to temporarily exceed its normal power limit when the CPU is using less than its allocated power. This parameter change is controlled through a driver that receives information about available power from the CPU and adjusts the GPU's power limit accordingly, enabling the GPU to operate at higher performance levels when power is available.
2Ease of operation
If separate power controllers are used for CPU and GPU to simplify power management, then each processor can be managed independently, but this prevents efficient utilization of the overall power budget when one processor is not using its full allocation
Solution Approach 1:
The patent merges the power management of the CPU and GPU at the software/driver level while maintaining separate hardware controllers. The driver combines information about CPU power consumption and available power budget to determine how much power can be allocated to the GPU, effectively creating a unified power management approach that leverages the simplicity of separate controllers while achieving efficient power utilization.
Solution Approach 2:
The driver acts as an intermediary between the CPU and GPU power management systems. It receives power consumption data from the CPU, calculates available power based on the overall power budget, and transmits this information to the GPU driver to adjust the GPU's power limit. This intermediary mechanism enables efficient power allocation without requiring complex hardware integration.
3Reliability
If the GPU operates at maximum power limit to ensure sufficient power for graphics operations, then graphics performance is maintained, but the overall system performance is hindered when the CPU is not utilizing its full power allocation
Solution Approach 1:
The GPU's power limit is made dynamic rather than static. Instead of always operating at a fixed maximum power limit, the GPU's power allocation adjusts in real-time based on the CPU's actual power consumption. This allows the GPU to scale its power usage and performance according to the available power budget, maximizing overall system performance while maintaining graphics operation stability.
Data Source
AI summary
A method for managing power in a multiple processor computing device includes detecting a first amount of power being used by a first processor of the computing device; determining an amount of extra power available based on the first amount of power and a power budget for the first processor; and transmits a value to a driver associated with a second processor of the computing device, wherein the value indicates the amount of extra power available, wherein the driver adjusts at least one operating parameter of the second processor based on the amount of extra power available.