Hierarchical Memory Packing Using Sensitivity-Based Precision

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Current computer systems face challenges in reducing data movement between levels of hierarchical memory systems, leading to increased access times and power consumption, as they often rely on moving large amounts of data across different memory levels, which is inefficient and costly.

Innovation Solution

The implementation of systems and methods that utilize sensitivity analysis and Massively Asymmetric Precision (MAP) data representations to reduce the number of bits transferred between memory hierarchies by identifying and eliminating 'garbage' bits in numerical data, allowing for more flexible and efficient data encoding and storage formats.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Quantity of substance

If data is stored in hierarchical memory structures with multiple levels, then storage capacity is increased, but data access time increases and power consumption increases due to data movement between levels

Engineering Contradiction:

Improvestorage capacityVSAvoiddata access time

Core Design Contradiction:

Quantity of substanceVSLoss of time

Solution Approach 1:

The patent extracts and eliminates 'garbage bits' from numerical data representations. By identifying and removing unnecessary bits that do not contribute to computational accuracy, the system reduces the total data volume that needs to be moved between memory levels, thereby reducing access time while maintaining storage capacity.

Inventive Principle:

Principle #2Taking out (Extraction)

Solution Approach 2:

The patent changes the precision parameter of data representations dynamically. By using sensitivity analysis to determine the actual precision requirements for different computations, the system adjusts the number of bits used to represent numerical data, storing only the necessary precision level in hierarchical memory structures.

Inventive Principle:

Principle #35Parameter changes

2Quantity of substance

If data is moved between levels of hierarchical memory systems, then storage capacity is utilized, but power consumption increases

Engineering Contradiction:

Improvestorage capacityVSAvoidpower consumption

Core Design Contradiction:

Quantity of substanceVSUse of energy by moving object

Solution Approach 1:

Inventive Principle:

Principle #2Taking out (Extraction)

Solution Approach 2:

Inventive Principle:

Principle #35Parameter changes

3Measurement precision

If standard precision data representations are used, then computational accuracy is maintained, but data movement volume increases

Engineering Contradiction:

Improvecomputational accuracyVSAvoiddata movement volume

Core Design Contradiction:

Measurement precisionVSQuantity of substance

Solution Approach 1:

The patent extracts and eliminates 'garbage bits' from numerical data representations. By identifying and removing unnecessary bits that do not contribute to computational accuracy, the system reduces data movement volume while preserving the precision actually required for correct computational results.

Inventive Principle:

Principle #2Taking out (Extraction)

Solution Approach 2:

The patent changes the precision parameter of data representations dynamically based on sensitivity analysis. By adjusting the number of bits used to represent numerical data according to the actual computational requirements, the system maintains computational accuracy while minimizing data movement volume.

Inventive Principle:

Principle #35Parameter changes

4Measurement precision

If more bits are used for data representation, then computational accuracy is improved, but memory capacity requirements increase

Engineering Contradiction:

Improvecomputational accuracyVSAvoidmemory capacity

Core Design Contradiction:

Measurement precisionVSQuantity of substance

Solution Approach 1:

The patent changes the precision parameter of data representations dynamically based on sensitivity analysis. By adjusting the number of bits used to represent numerical data according to the actual computational requirements, the system achieves the necessary computational accuracy while minimizing memory capacity requirements.

Inventive Principle:

Principle #35Parameter changes

Data Source

PatentUS10114554B1Arrangements for storing more data in faster memory when using a hierarchical memory structure

Publication Date: 2018.10.30 INTELLECTUAL PROPERTY SYSTEMS LLC

AI summary

Data employed in computations is processed so that during computations more of the data can be fit into or maintained in a smaller but higher speed memory than an original source of the data. More specifically, a sensitivity value is determined for various items of the data which reflect the number of bits in the data items that are not garbage bits, and only information in the data items that are indicated by the sensitivity value to not be garbage bits are necessarily effectively retained. At least the information that is not garbage bits and the corresponding associated sensitivity are packed together. The results of computations that are performed using the data items as at least one of the operands for the computation are associated with a sensitivity that is derived from the individual sensitivities of the operands used in the computation.