Nonvolatile Memory Write Timing for Host Response Deadlines

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Memory systems with nonvolatile memory face delays in writing and responding to write requests, leading to potential failure in meeting the host's expected response time due to delayed write completion mechanisms.

Innovation Solution

Implementing a controller that initiates a write process after a predetermined time lapse from the write request and ensures a response within a designated time limit by triggering a forced response when the elapsed time exceeds a threshold, using parameters like forced_completion_time and process_time to manage write operations.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If delayed write completion is implemented to efficiently use storage area, then storage efficiency is improved, but response time to host is worsened

Engineering Contradiction:
Improvestorage efficiencyVSAvoidresponse time
Core Design Contradiction:
ProductivityVSLoss of time

Solution Approach 1:

The controller performs preliminary actions by buffering write requests and proactively managing write operations before the host times out. The controller monitors elapsed time since write requests and initiates writes to the nonvolatile memory in advance, ensuring completion before the host's expected response time expires. This preliminary action prevents the host from perceiving delays while maintaining efficient bulk write operations.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The system dynamically adjusts write operation timing based on multiple factors including elapsed time since write requests, buffer status, and host expectations. The controller transitions between immediate write execution and delayed write completion based on real-time conditions, optimizing both storage efficiency and response time by making dynamic decisions about when to initiate write operations to the nonvolatile memory.

Inventive Principle:
Principle #15Dynamics

2Productivity

If writing is delayed until multiple write requests are received, then storage area usage is improved, but write completion timing is worsened

Engineering Contradiction:
Improvestorage area usageVSAvoidwrite completion timing
Core Design Contradiction:
ProductivityVSManufacturing precision

Solution Approach 1:

The controller implements feedback mechanisms by continuously monitoring elapsed time since write requests are received and buffer status. Based on this feedback, the controller decides whether to immediately execute writes or delay them for bulk processing. The system adjusts write timing dynamically, using feedback from time monitoring and buffer status to maintain precise write completion timing while optimizing storage area utilization.

Inventive Principle:
Principle #23Feedback

Solution Approach 2:

The system changes operational parameters by adjusting write timing based on elapsed time thresholds and buffer fullness. When the buffer reaches a certain fullness threshold or when elapsed time approaches the host's expected response time, the system transitions from delayed write completion to immediate write execution. This parameter-based control allows flexible adjustment between storage efficiency and write completion precision.

Inventive Principle:
Principle #35Parameter changes

Data Source

PatentUS20250251873A1Memory system
Publication Date: 2025.08.07 KIOXIA CORP
  • US20250251873A1 patent drawing
  • US20250251873A1 patent drawing
  • US20250251873A1 patent drawing

AI summary

According to one embodiment, a memory system includes a nonvolatile memory and a controller. The controller receives a first write request associated with first data from a host. In response to a lapse of first time since the reception of the first write request, the controller starts a write process of second data to the nonvolatile memory. The second data includes at least the first data. The controller transmits a first response to the first write request to the host in response to completion of the write process. The first time is time obtained by subtracting second time from third time designated by the host as a time limit of the transmission of the first response since the reception of the first write request.