Memory Bus Data Consolidation for Power Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current memory systems face challenges in reducing power consumption, as voltage scaling has reached a plateau, and the next generation of low-power memory standards offers only marginal power savings, necessitating alternative methods to decrease energy usage.
Innovation Solution
The implementation of data consolidation techniques in memory systems, where repetitive data is sent only once and replicated at the receiving end, reducing the amount of data transferred across the memory bus, thereby minimizing power consumption.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Use of energy by moving object
If voltage scaling is used to reduce power consumption, then power consumption decreases, but voltage scaling has reached a plateau and offers only marginal power savings
Solution Approach 1:
The patent applies the copying principle by transmitting a single instance of repetitive data over the memory bus and using a data mask lane to indicate which data elements should be replicated at the receiver end. Instead of sending multiple copies of identical data across multiple lanes, the system sends one copy and uses the DM lane to convey replication instructions, thereby reducing bus traffic while maintaining data delivery completeness.
Solution Approach 2:
The patent extracts the repetitive data pattern recognition function from the traditional memory bus protocol and implements it through a dedicated data mask lane. The DM lane is extracted as a separate control channel that carries information about which data elements need replication, separating the control signaling from the data transmission itself and enabling more efficient power consumption.
2Reliability
If data is sent repeatedly across multiple data lanes, then data transfer completeness is ensured, but power consumption increases
Solution Approach 1:
The patent applies the copying principle by transmitting a single instance of repetitive data over the memory bus and using a data mask lane to indicate which data elements should be replicated at the receiver end. Instead of sending multiple copies of identical data across multiple lanes, the system sends one copy and uses the DM lane to convey replication instructions, thereby reducing bus traffic while maintaining data delivery completeness.
Solution Approach 2:
The data mask lane serves as an intermediary control channel that mediates between the data to be transmitted and the replication process. It carries control information about which data elements should be replicated without carrying the actual data payload, enabling the system to achieve complete data transfer with reduced power consumption by separating control signaling from data transmission.
3Speed
If all data lanes are activated for data transfer, then data transfer speed is maximized, but power consumption increases
Solution Approach 1:
The patent applies partial action by activating only the necessary subset of data lanes for transmitting unique data elements, rather than activating all available lanes. The data mask lane indicates which specific lanes should remain active and which can be deactivated, enabling the system to achieve sufficient data transfer speed while minimizing the number of active lanes and reducing power consumption proportionally.
Solution Approach 2:
The system dynamically adjusts the number of active data lanes based on the actual data patterns being transmitted. By using the data mask lane to convey information about data repetitiveness, the system can dynamically deactivate unnecessary lanes during transmission, optimizing the balance between transfer speed and power consumption for each specific data transfer operation.
Data Source
AI summary
Power saving techniques for memory systems are disclosed. In particular, exemplary aspects of the present disclosure contemplate taking advantage of patterns that may exist within memory elements and eliminating duplicative data transfers. Specifically, if data is repetitive, instead of sending the same data repeatedly, the data may be sent only a single time with instructions that cause the data to be replicated at a receiving end to restore the data to its original repeated state.


