Namespace L2P Mapping Granularity for Mixed-Write Memory Subsystems

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing memory sub-systems inefficiently manage namespace mapping, leading to suboptimal allocation of memory resources for namespace L2P tables, affecting performance and cost efficiency, particularly in scenarios with both sequential and random writes.

Innovation Solution

Adaptive memory allocation for namespace L2P tables based on parameters such as namespace capacity and granularity, dynamically adjusting memory usage according to the write patterns of each namespace.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If memory is allocated for namespace L2P tables based on maximum capacity, then all namespaces can be supported, but memory resources are wasted when namespaces are not fully utilized

Engineering Contradiction:

Improvenamespace support capabilityVSAvoidmemory resource allocation

Core Design Contradiction:

Adaptability or versatilityVSQuantity of substance

Solution Approach 1:

The patent implements dynamic memory allocation for namespace L2P tables by introducing a granularity parameter that allows the memory allocation to adapt to the actual usage patterns of each namespace. Instead of static allocation based on maximum capacity, the system dynamically adjusts memory allocation according to the granularity setting, which determines the mapping resolution between namespace logical addresses and device physical addresses. This enables the memory allocation to scale with actual namespace utilization rather than predetermined maximums.

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

The patent changes the allocation parameters from fixed maximum capacity to variable granularity-based allocation. By introducing granularity as a configurable parameter (e.g., 128 KB, 256 KB, 512 KB), the system can adjust the memory allocation granularity to match the actual namespace requirements. This parameter change allows fine-tuned control over memory usage, allocating more memory only when higher granularity is needed for specific namespaces while using less memory for namespaces that can operate with coarser granularity.

Inventive Principle:

Principle #35Parameter changes

2Productivity

If finer granularity is used for namespace L2P tables, then random write performance improves, but memory consumption increases

Engineering Contradiction:

Improverandom write performanceVSAvoidmemory consumption

Core Design Contradiction:

ProductivityVSQuantity of substance

Solution Approach 1:

The patent applies local quality by allowing different granularity settings for different namespaces based on their specific access patterns and performance requirements. Instead of using a uniform fine granularity across all namespaces, the system configures granularity locally for each namespace according to its workload characteristics. Namespaces with random write workloads can use finer granularity to improve performance, while namespaces with sequential access patterns can use coarser granularity to reduce memory consumption.

Inventive Principle:

Principle #3Local quality

3Adaptability or versatility

If more memory is allocated for namespace metadata, then namespace management flexibility increases, but cost efficiency decreases

Engineering Contradiction:

Improvenamespace management flexibilityVSAvoidcost efficiency

Core Design Contradiction:

Adaptability or versatilityVSEase of manufacture

Solution Approach 1:

The patent implements partial action by allocating memory resources partially based on actual namespace requirements rather than providing full memory allocation for all possible namespaces. The granularity-based allocation ensures that memory is allocated only to the extent needed for each namespace's operational requirements, avoiding excessive memory provisioning. This partial allocation approach maintains sufficient namespace management flexibility while reducing overall memory consumption and associated costs.

Inventive Principle:

Principle #16Partial or excessive action

Data Source

PatentUS12602158B2Managing namespace mapping in a memory sub-system

Publication Date: 2026.04.14 MICRON TECHNOLOGY INC

AI summary

A system includes a first memory device; a second memory device; and a processing device, operatively coupled with the first memory device and the second memory device, to perform operations including: determining a size of the first memory device designated for storing a plurality of address mapping data structures, wherein each address mapping data structure of the plurality of address mapping data structures is associated with a corresponding namespace of a plurality of namespaces, wherein each namespace of the plurality of namespaces is associated with a corresponding region of a second memory device; associating a first address mapping granularity level with the first memory device for storing a first address mapping data structure of the plurality of address mapping data structures; determining a remaining available space of the first memory device by excluding a first size of the first address mapping data structure from the size of the first memory device, wherein the first size is calculated based on the first address mapping granularity level; and associating, in view of the remaining available space of the first memory device and a remaining available space of the second memory device, a second address mapping granularity level with the first memory device for storing a second address mapping data structure of the plurality of address mapping data structures.