Logical-to-Physical Bank Mapping Cache for Quad-Port Memory
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current memory devices struggle to support high-bandwidth, on-chip memory that can concurrently handle multiple read and write accesses with low latency, leading to inefficiencies in area and power consumption, especially in custom-designed dual-port memory solutions.
Innovation Solution
A memory device with a logical-to-physical bank mapping cache that allows up to two read and two write operations in a single clock cycle, utilizing a plurality of physical memory banks and a cache to manage logical memory banks, enabling quad-port access while reducing power and area consumption compared to custom-built dual-pumped dual-port memory.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If custom-designed dual-port memory is used to support multiple simultaneous read and write operations, then the memory can handle concurrent accesses, but the area and power consumption increase significantly
Solution Approach 1:
The memory system is segmented into multiple independent memory banks (first bank, second bank, third bank, fourth bank) that can operate simultaneously. Each bank handles specific access patterns, allowing parallel read and write operations without requiring a fully custom dual-port design. This segmentation enables concurrent accesses while maintaining area efficiency by using standard memory cell structures.
Solution Approach 2:
The memory device provides universal quad-port access capability through a standardized architecture that can handle various access patterns (read-read, write-write, read-write combinations) across multiple banks. The control logic universally manages address routing and data flow to support up to two simultaneous read operations and two simultaneous write operations without requiring custom design for each specific access pattern.
2Productivity
If custom-designed dual-port memory is used to support multiple simultaneous read and write operations, then the memory can handle concurrent accesses, but the power consumption increases significantly
Solution Approach 1:
By segmenting the memory into multiple banks that can be independently activated, the system only powers the specific banks needed for current operations. This selective activation reduces overall power consumption compared to a custom dual-port design where all memory circuits would be actively managed regardless of usage pattern.
Solution Approach 2:
The control logic dynamically routes addresses and activates only the necessary memory banks based on the current access pattern. This dynamic bank selection and address routing minimizes the number of simultaneously active memory circuits, thereby reducing power consumption while maintaining the ability to support concurrent read and write operations.
3Productivity
If single-port SRAM memory runs at twice the clock frequency to support two simultaneous operations, then dual-port access is achieved, but latency increases
Solution Approach 1:
The memory is divided into multiple banks that can be accessed simultaneously by different ports. This allows true parallel access where each bank handles its own operations independently, eliminating the need to run a single bank at doubled frequency. Consequently, access latency is reduced because operations occur in parallel across banks rather than sequentially within a single high-speed bank.
Solution Approach 2:
The patent transitions from a single-dimension approach (one bank running at 2x frequency) to a multi-dimensional approach (multiple banks running at normal frequency with parallel access). By adding the bank dimension for parallelism, the system achieves dual-port functionality without the latency penalty of doubled clock frequency operation.
4Productivity
If 4-port register files are used to allow two read and two write operations simultaneously, then high concurrency is achieved, but area density becomes 3 times worse
Solution Approach 1:
The memory device segments functionality across four separate memory banks, each optimized for standard density. This segmentation allows the system to achieve 4-port equivalent functionality (two simultaneous reads and two simultaneous writes) using conventional memory cell structures rather than specialized high-speed register file cells, thereby maintaining area density comparable to standard SRAM.
Solution Approach 2:
The control logic provides universal quad-port access capability by dynamically routing addresses and data flows across the four banks. This standardized multi-functional approach achieves the same concurrency as custom 4-port register files but with better area density by using standard memory structures rather than specialized high-speed cells.
Data Source
AI summary
A memory device with a logical-to-physical (LTP) bank mapping cache that supports multiple read and write accesses is described herein. The memory device allows for at least one read operation and one write operation to be received during the same clock cycle. In the event that the incoming write operation is not blocked by the at least one read operation, data for that incoming write operation may be stored in the physical memory bank corresponding to a logical memory bank that is associated with the incoming write operation. In the event that the incoming write operation is blocked by the at least one read operation, then data for that incoming write operation may be stored in an unmapped physical bank that is not associated with any logical memory bank.


