Floating-Point Normalization Logic with Parallel Correction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing data processing techniques for normalizing floating-point numbers suffer from timing delays due to the need for correction multiplexers and increased complexity, particularly in predicting the shift count for placing the leading one in the correct bit position, which complicates the design and increases costs.
Innovation Solution
A data processing apparatus with prediction logic to generate a shift indication, correction logic to determine if the prediction is correct or incorrect based on the least significant bit of the shift indication, and shift logic to apply a correction shift if necessary, allowing for parallel operation and reducing complexity and timing delays.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If leading one prediction logic is used to predict the shift count, then the normalization process can proceed in parallel, but the prediction may be incorrect by one bit position causing timing delays
Solution Approach 1:
The patent applies preliminary action by performing the correction shift operation in advance, parallel to the normalization shift logic. The correction logic evaluates whether a correction is needed based on the least significant bit of the shift indication and the data value, and the correction shift logic is ready to apply the correction immediately without waiting for the normalization shift logic to complete, thus eliminating timing delays
Solution Approach 2:
The patent introduces an intermediary correction logic unit that acts as a mediator between the leading one prediction logic and the final normalized result. This correction logic evaluates the prediction accuracy by examining the least significant bit of the shift indication and the data value, and generates an output signal that controls the correction shift logic to apply the necessary correction, thus resolving the timing conflict
2Measurement precision
If correction logic is added to detect prediction errors, then accuracy is improved, but device complexity increases
Solution Approach 1:
The patent extracts only the essential information needed for correction by focusing solely on the least significant bit of the shift indication and the corresponding bit of the data value. This selective extraction approach enables accurate prediction error detection without requiring complex analysis of the entire shift count or data value, thus maintaining simplicity while improving accuracy
Solution Approach 2:
The patent changes the parameter being monitored from the entire shift count value to just the least significant bit of the shift indication. This parameter change simplifies the correction logic significantly, as it only needs to evaluate a single bit position rather than the complete shift count, reducing device complexity while maintaining the ability to detect prediction errors
Data Source
AI summary
A data processing apparatus and method are provided for normalizing a data value to produce a result value. The data processing apparatus includes prediction logic for generating a shift indication based on a prediction of the number of bit positions by which the data value needs to be shifted in order to normalize the data value. Further, normalizer logic is used to apply a shift operation to the data value based on the shift indication. In addition, correction logic is operable in parallel with the normalizer logic to determine from the data value and a least significant bit of the shift indication whether the shift indication has correctly predicted the number of bit positions by which the data value needs to be shifted in order to normalize the data value, or whether instead the prediction is incorrect, and to generate an output signal dependent on that determination. Shift logic is then used, if the output signal indicates that the prediction is incorrect, to apply a correction shift such that the result value is the normalized data value.