Dynamic Peak Power Management for Multi-Die Memory Subsystems
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In memory subsystems, peak power operations by multiple memory components can lead to voltage drops and error-prone conditions due to high current demands, necessitating conservative power management that limits concurrent peak power operations to avoid cumulative power consumption exceeding limits, especially in mobile systems.
Innovation Solution
Memory components communicate their activity states to dynamically determine the number of active components, allowing only a threshold to perform peak power operations, thereby optimizing power consumption and enabling more components to execute peak operations when others are inactive.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conservative power management limits concurrent peak power operations, then voltage drops and error-prone conditions are avoided, but system performance deteriorates due to fewer components performing peak operations
Solution Approach 1:
The patent implements dynamic power management where the threshold for allowing peak power operations is adjusted based on real-time power network conditions. The power manager continuously monitors power consumption and dynamically modifies which memory components can perform peak power operations, transitioning from static conservative limits to adaptive dynamic control that optimizes both reliability and performance
Solution Approach 2:
The system changes operational parameters by adjusting the threshold number of components allowed to perform peak power operations based on monitored power consumption levels. When power consumption is below thresholds, more components are permitted to perform peak operations; when thresholds are approached, the limit is reduced, thereby adapting system behavior to current power conditions
2Use of energy by stationary object
If all dice are assumed active consuming worst-case power, then power consumption limit is maintained, but opportunities are forgone to allow more dice to perform peak operations when others are inactive
Solution Approach 1:
The patent implements feedback mechanisms where memory components communicate their actual activity state to the power manager. This feedback allows the system to distinguish between active and inactive components, replacing the worst-case assumption with real-time information about actual power consumption patterns, enabling more accurate power management decisions
Solution Approach 2:
Memory components self-report their activity state to the power manager, eliminating the need for the power manager to assume worst-case scenarios. Each component provides information about its own power consumption status, allowing the system to make informed decisions about allowing peak power operations without conservative overestimation
Data Source
AI summary
Exemplary methods, apparatuses, and systems include a first die in a power network receiving, from each die of a plurality of dice in the power network, a first signal indicating that the respective die of the plurality of dice is in a high current state or a second signal indicating that the respective die of the plurality of dice is an active current state. The received signals include at least one second signal. The first die determines, based upon the received signals, a number of dice of the plurality of dice that are currently active and selects an activity threshold based upon that number. The first die further determines an activity level for the power network and transmits, to the plurality of dice, the first signal indicating that the first die is in the high current state in response to determining that the activity level is less than the activity threshold.