Probe Storage Memory Random Access Performance

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Probe storage memory devices face limitations in random read/write performance and reliability due to the need for all probes to move together, restricting the number of simultaneous data operations and leading to idle probe tips and increased seek times, while high parallelism benefits sequential data access but not random access.

Innovation Solution

Implementing a hierarchical organization of probe grouping and clustering to maximize parallelism, subdividing tracks into smaller data zones, and splitting the traditional single mover architecture into multiple movers to enable independent motion and overlapping read/write operations, along with using a non-volatile semiconductor cache for improved performance.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If all probes move together in a single mover architecture, then sequential data access benefits from high parallelism, but random read/write performance deteriorates due to restricted simultaneous operations and increased seek times

Engineering Contradiction:
Improvesequential data access throughputVSAvoidseek time
Core Design Contradiction:
ProductivityVSLoss of time

Solution Approach 1:

The patent divides the storage medium into multiple independent movers, each capable of independent motion. This segmentation allows different groups of probes to access different data zones simultaneously, enabling random access operations to occur in parallel without requiring all probes to travel together, thus reducing seek time while maintaining sequential access efficiency.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent introduces a new dimension of organization by grouping probes into multiple movers that can operate independently in the spatial dimension. This allows the system to transition from a single-mover sequential access model to a multi-mover parallel access model, where probes can be directed to different movers based on data location, improving random access performance.

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

2Device complexity

If probes are organized in a single mover, then device complexity is low, but random IOPS performance deteriorates due to idle probe tips and limited simultaneous operations

Engineering Contradiction:
Improvemover architecture complexityVSAvoidrandom IOPS
Core Design Contradiction:
Device complexityVSProductivity

Solution Approach 1:

The storage system is segmented into multiple movers, each handling specific data zones. This allows probes to be actively engaged on multiple movers simultaneously, increasing random IOPS by eliminating idle probe tips while distributing the control complexity across multiple independent units rather than one complex centralized system.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent implements dynamic assignment of probes to movers based on access patterns and data location. This dynamic organization allows the system to optimize probe utilization for random access operations, ensuring that probes are actively working on multiple movers simultaneously rather than remaining idle, thereby improving random IOPS performance.

Inventive Principle:
Principle #15Dynamics

3Quantity of substance

If track density is increased to achieve high capacity, then storage capacity improves, but reliability deteriorates due to increased susceptibility to errors and reduced margin for probe positioning accuracy

Engineering Contradiction:
Improvestorage capacityVSAvoiddata access reliability
Core Design Contradiction:
Quantity of substanceVSReliability

Solution Approach 1:

By dividing the high-density storage medium into multiple movers, the patent reduces the physical distance probes must travel to access data within each mover's zone. This segmentation maintains the high overall capacity while improving reliability by reducing positioning errors and susceptibility to disturbances during data access operations.

Inventive Principle:
Principle #1Segmentation

Applied Scientific Principles

This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.

Function Achieved in This Case

Enhances random IOPS by allowing more probes to cooperate in data operations, reduces seek and scan times, and increases data throughput while maintaining high capacity and reducing probe tip wear, with potential for improved power efficiency and cost-effectiveness.

Implementation Method 1

storage media made of materials that can be electrically switched between two or more states having different electrical characteristics such as resistance, polarization dipole direction

Methodology Applied
Scientific EffectElectrical switching between states:

Implementation Method 2

Passing a current through the storage media, or applying an electric field to the media, is typically accomplished by applying a voltage between a sharp probe tip on one side of the storage media and an electrode on the other side of the storage media

Methodology Applied
Scientific EffectElectric field application: Electric Field

Data Source

PatentUS8149680B2Random read/write performance of probe storage memory devices
Publication Date: 2012.04.03 INTEL CORP
  • US8149680B2 patent drawing
  • US8149680B2 patent drawing
  • US8149680B2 patent drawing

AI summary

Current probe-type memory architecture assumes that the minimum chunk of data that a probe tip can access is one entire track and perhaps only four out of five-thousand, for example, probes participate in the access thereby degrading performance. By subdividing the track into D finer chunks or data zones, D times more probes can cooperate to read out the data, hence increasing the data throughput by Dx. Each tip now only scans approximately one Dth of the track and hence the scan time is reduced by a factor D, while D probes are being utilized in parallel.