Multi-Plane Cache Transfer Using Offset Latch Timing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current memory systems experience delays in multi-plane operations due to wait times between the transmission of data pages, which hinder faster data transfers and higher throughput.
Innovation Solution
Data portions are transferred between subsets of latches at offsetting times, allowing concurrent receipt or transmission of pages without delay during write or read operations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If data pages are transmitted sequentially in multi-plane operations, then data transfer reliability is maintained, but wait times between pages increase and throughput decreases
Solution Approach 1:
The patent segments the multi-plane operation into multiple independent data transfer channels, where each plane can transfer data pages concurrently without waiting for others. This segmentation allows parallel processing of data pages across different planes, eliminating sequential wait times and improving overall throughput while maintaining transfer reliability through independent error handling for each segment.
Solution Approach 2:
The patent implements preliminary actions by pre-configuring multiple planes and their associated latches before data transfer begins. This allows the system to have multiple ready-to-transfer channels established in advance, enabling immediate concurrent data page transfers without setup delays between pages, thus reducing wait time and improving throughput.
2Speed
If wait times are introduced between data page transmissions, then data transfer accuracy is ensured, but transfer speed decreases
Solution Approach 1:
The patent achieves continuity of useful action by maintaining active data transfer operations across multiple planes simultaneously. Instead of interrupting transfer to introduce wait times between pages, the system continues useful action through parallel transfers on multiple planes, where the completion of one plane's transfer does not halt others. This continuous operation increases data transfer speed while eliminating idle wait time.
Solution Approach 2:
By segmenting the data transfer into multiple independent plane operations, the patent allows each segment to proceed at full speed without being bottlenecked by wait times required for sequential processing. Each segmented plane transfer maintains accuracy through its own error checking while proceeding concurrently, thus improving overall transfer speed without sacrificing data integrity.
3Productivity
If multiple planes operate concurrently, then throughput increases, but system complexity increases
Solution Approach 1:
The patent manages multi-plane complexity through systematic segmentation, where each plane is divided into independent, identically structured units with dedicated latches and transfer logic. This modular segmentation allows concurrent operation of multiple planes while keeping individual plane complexity manageable and reusable across all planes, thus enabling high throughput without proportionally increasing overall system complexity.
Solution Approach 2:
The patent applies universality by designing identical, reusable plane structures that can be replicated across multiple planes. Each plane uses the same latch configuration and transfer mechanism, allowing the system to achieve high throughput through concurrency while minimizing complexity through design reuse. This universal structure reduces the need for unique control logic for each plane, managing system complexity effectively.
Data Source
AI summary
Methods, systems, and devices for multi-plane cache transfer enhancement are described. Portions of a page associated with different subsets of planes may be transferred between subsets of latches at offsetting times such that a portion of a page may be transferred concurrent with a portion of another page being received or transmitted. For a write operation, a portion of a page may be transferred from a gateway latch to a buffer latch concurrent with another portion of a page being externally received. For a read operation, a portion of a page may be transferred from a buffer latch to a gateway latch concurrent with another portion of a page being externally transmitted. This may allow the external receipt and transmission of pages to be performed with no delay between the pages.