Floating-Point Exceptional Value Encoding With Partial Exponent Read

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Conventional encoding of exceptional values in floating point number formats requires reading all exponent bits, leading to increased circuit complexity, power consumption, and delay in floating point unit (FPU) operations.

Innovation Solution

A normalized floating point format is introduced, where exceptional values are encoded using a subset of exponent bits within an unused range, allowing detection without reading all exponent bits, and optionally using a further portion of exponent bits for type indication.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If conventional encoding of exceptional values is used (requiring all exponent bits to be read), then reliable detection of exceptional values is achieved, but circuit complexity, power consumption, and operation delay increase

Engineering Contradiction:

Improvedetection reliabilityVSAvoidcircuit complexity

Core Design Contradiction:

ReliabilityVSDevice complexity

Solution Approach 1:

The patent extracts only the necessary subset of exponent bits (e.g., most significant bits) required to detect exceptional values, rather than reading all exponent bits. This extraction approach reduces the number of bits that need to be read and processed, directly lowering circuit complexity and power consumption while maintaining reliable detection of exceptional values such as infinities and NaNs.

Inventive Principle:

Principle #2Taking out (Extraction)

Solution Approach 2:

The patent applies partial action by reading only a portion of the exponent bits (partial reading) instead of the complete exponent field. By reading fewer bits than the total available, the system achieves the minimum necessary action for reliable detection while avoiding the excessive resource consumption associated with reading all bits, thus reducing circuit complexity and operation delay.

Inventive Principle:

Principle #16Partial or excessive action

2Loss of information

If conventional encoding of exceptional values is used (requiring all exponent bits to be read), then complete information about exceptional values is obtained, but power consumption and operation delay increase

Engineering Contradiction:

Improveinformation completenessVSAvoidpower consumption

Core Design Contradiction:

Loss of informationVSUse of energy by stationary object

Solution Approach 1:

The patent extracts only the essential information from the exponent field by reading a subset of bits sufficient to identify exceptional values. This selective extraction maintains the necessary information completeness for detecting infinities and NaNs while significantly reducing the energy required to read and process the exponent field compared to reading all bits.

Inventive Principle:

Principle #2Taking out (Extraction)

Solution Approach 2:

The patent performs partial reading of the exponent bits, applying partial action to obtain only the necessary information for exceptional value detection. This approach prevents excessive energy consumption associated with reading the entire exponent field while ensuring that no critical information is lost for reliable detection.

Inventive Principle:

Principle #16Partial or excessive action

3Measurement precision

If conventional encoding of exceptional values is used (requiring all exponent bits to be read), then accurate classification of exceptional value types is achieved, but operation delay increases

Engineering Contradiction:

Improveclassification precisionVSAvoidoperation delay

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

The patent extracts the most significant bits of the exponent field that contain the critical information for classifying exceptional value types. By focusing on these extracted bits rather than processing the entire exponent field, the system achieves accurate classification of exceptional values (distinguishing between infinities, NaNs, and normal values) while reducing the time required for detection and classification operations.

Inventive Principle:

Principle #2Taking out (Extraction)

4Device complexity

If normalized floating point format with subset encoding is used, then circuit complexity and power consumption are reduced, but encoding precision for exceptional values may be affected

Engineering Contradiction:

Improvecircuit complexityVSAvoidencoding precision

Core Design Contradiction:

Device complexityVSManufacturing precision

Solution Approach 1:

The patent applies partial reading action to the exponent field, reading only the necessary subset of bits for exceptional value detection. This partial action approach reduces circuit complexity and power consumption while maintaining sufficient encoding precision by ensuring that the read subset contains all the information needed to accurately distinguish between different exceptional value types and normal values.

Inventive Principle:

Principle #16Partial or excessive action

Data Source

PatentUS20260072641A1Encoding Exceptional Values in a Floating Point Number Format

Publication Date: 2026.03.12 IMAGINATION TECH LTD

AI summary

A processor is able to re-encode a normal value of a first floating point format to a normalized floating point format, and re-encode a subnormal value of the first floating point format to the normalized floating point format. A subset of the exponent bits of the exponent field of a normalized floating point format is used to encode a predetermined indication of being an exceptional value, wherein a combination of bit values used to encode this predetermined indication does not occur within a range of the exponent field of the normalized floating point format used to represent any possible non-exceptional, normal and subnormal values.