Stacked Memory Die Access Using Master-Slave Data Combining
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current memory systems face challenges in achieving high bandwidth without increasing pin count or clock speed, which can lead to increased costs and complexity, especially in mobile systems where power efficiency and compact design are crucial.
Innovation Solution
The use of stacked memory dice with a master-slave configuration, where one memory die acts as a master and another as a slave, allows for simultaneous prefetching and data transfer without additional pins or increased clock speed, utilizing modulation schemes like PAM4 to enhance data transfer efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If separate command and data channels are used for memory access, then control and data transfer are independent, but access time increases due to sequential operation
Solution Approach 1:
The patent combines separate command and data channels into a unified stacked memory dice architecture where command and data paths are integrated through shared interconnect structures. Multiple memory dice are stacked with interconnect layers that enable simultaneous command routing and data transfer across dice boundaries, allowing parallel operation of previously sequential channels.
Solution Approach 2:
The patent transitions from planar 2D memory organization to 3D stacked architecture. By stacking memory dice vertically and implementing interconnect layers between dice, the system adds a vertical dimension to data and command paths. This enables multiple command and data channels to operate simultaneously in different spatial layers, reducing access time while maintaining control independence through dedicated routing paths.
2Quantity of substance
If multiple memory dice are stacked to increase capacity, then storage density increases, but access complexity increases due to multiple access paths
Solution Approach 1:
The patent segments the memory system into discrete dice stacked vertically, with each die containing independent memory arrays. Command and data paths are segmented into dedicated channels that route through specific interconnect layers. This segmentation allows the controller to address individual dice or groups of dice independently, managing access complexity through modular organization rather than monolithic structure.
Solution Approach 2:
The patent introduces interconnect layers as intermediary structures between stacked memory dice. These interconnect layers provide standardized routing interfaces that simplify complex inter-die communication. The interconnects act as mediators that manage signal routing, timing, and protocol conversion between dice, reducing access complexity by abstracting the underlying physical complexity into standardized interface operations.
3Productivity
If read and write operations share the same memory channels, then channel utilization is maximized, but operation conflicts increase leading to reduced throughput
Solution Approach 1:
The patent implements dynamic channel allocation where command and data paths can be dynamically assigned to different memory dice based on operation type and address. The interconnect structure enables flexible routing that adapts to read/write operation requirements in real-time. This dynamic routing reduces operation conflicts by directing simultaneous reads and writes through different physical paths while maintaining high channel utilization through efficient resource sharing.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Methods, systems, and devices for stacked memory dice and combined access operations are described. A device may include multiple memory dice. One die may be configured as a master, and another may be configured as a slave. The master may communicate with a host device. A slave may be coupled with the master but not the host device. The device may include a first die (e.g., master) and a second die (e.g., slave ). The first die may be coupled with a host device and configured to output a set of data in response to a read command. The first die may supply a first subset of the data and obtain a second subset of the data from the second die. In some cases, the first die may select, based on a data rate, a modulation scheme (e.g., PAM4, NRZ, etc.) and output the data using the selected modulation scheme.