Processor Trivial Operand Handling for Sparse Matrix Efficiency

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing processor technologies incur significant computational overhead due to the performance of complex arithmetic operations on trivial operands, such as multiplication by 0, +1, or -1, especially in applications involving sparse matrices, where many operations are trivial and current methods only halt complex operations rather than eliminating them.

Innovation Solution

Employing trivial operand flags to identify and manage trivial operands, such as 0, 1, and -1, within a processor's register file, allowing for parallel comparison and storage, thereby avoiding unnecessary complex calculations and reducing computational latency and power consumption.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If complex arithmetic operations are performed on all operands, then calculation accuracy is maintained, but computational time and power consumption increase significantly for trivial operations

Engineering Contradiction:
Improvecalculation accuracyVSAvoidcomputational time
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The patent applies preliminary action by performing a preliminary check to determine whether an operand is trivial before executing complex arithmetic operations. The processor checks if operands are trivial values (0, 1, or -1) and only performs complex operations when necessary, thereby avoiding unnecessary computational overhead while maintaining accuracy when required

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent changes the operational parameters dynamically based on the type of operand. When a trivial operand is detected, the processor switches from complex arithmetic operation mode to a simplified operation mode that directly produces the correct result without performing full multiplication or addition sequences, thus reducing computational time while maintaining accuracy

Inventive Principle:
Principle #35Parameter changes

2Measurement precision

If complex arithmetic operations are performed on all operands, then calculation accuracy is maintained, but power consumption increases significantly for trivial operations

Engineering Contradiction:
Improvecalculation accuracyVSAvoidpower consumption
Core Design Contradiction:
Measurement precisionVSUse of energy by moving object

Solution Approach 1:

The patent applies preliminary action by performing a preliminary check to determine whether an operand is trivial before executing complex arithmetic operations. The processor checks if operands are trivial values (0, 1, or -1) and only performs complex operations when necessary, thereby avoiding unnecessary computational overhead while maintaining accuracy when required

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent changes the operational parameters dynamically based on the type of operand. When a trivial operand is detected, the processor switches from complex arithmetic operation mode to a simplified operation mode that directly produces the correct result without performing full multiplication or addition sequences, thus reducing computational time while maintaining accuracy

Inventive Principle:
Principle #35Parameter changes

3Quantity of substance

If sparse matrices are stored in compressed format, then storage efficiency is improved, but address generation overhead increases and performance deteriorates

Engineering Contradiction:
Improvestorage efficiencyVSAvoidperformance
Core Design Contradiction:
Quantity of substanceVSProductivity

Solution Approach 1:

The patent applies self-service by enabling the processor to automatically detect and handle trivial operations within sparse matrix computations. The processor independently identifies trivial operands during computation and applies optimized handling without requiring complex address generation or external control, thereby improving performance while maintaining storage efficiency

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The patent changes the computational approach based on the detected operand type. When trivial operands are identified in sparse matrix operations, the processor applies simplified computation paths that avoid the overhead of compressed format address generation, thereby improving performance for the significant portion of operations involving trivial values

Inventive Principle:
Principle #35Parameter changes

Data Source

PatentEP3287894B1A processor exploiting trivial arithmetic operations
Publication Date: 2019.10.23 MOVIDIUS LTD
  • EP3287894B1 patent drawingFigure 1
  • EP3287894B1 patent drawingFigure 2

AI summary

The present application relates to the field of processors and in particular to the carrying out of arithmetic operations. Many of the computations performed by processors consist of a large number of simple operations. As a result, a multiplication operation may take a significant number of clock cycles to complete. The present application provides a processor having a trivial operand register, which is used in the carrying out of arithmetic or storage operations for data values stored in a data store.