Shared Special Case Handling Circuit for Multi-Lane Processing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Data processing systems with multiple lanes of processing circuitry face increased circuit area, cost, and complexity due to dedicated special case handling circuits for rare operand-dependent special case conditions, and software-based solutions are slow and inefficient.
Innovation Solution
Implementing shared special case handling circuits that serve multiple lanes of processing circuitry, detecting special case conditions in parallel and performing necessary operations in series, thereby reducing circuit area overhead while maintaining acceptable performance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If dedicated special case handling circuits are provided within each lane of processing circuitry, then special case processing operations can be performed correctly, but circuit area, cost, and complexity increase significantly
Solution Approach 1:
The patent merges the special case handling functionality from multiple lanes into a single shared circuit. The special case handling circuit is provided outside the parallel processing lanes and receives inputs from all lanes, consolidating what would otherwise be duplicate circuits in each lane. This merging reduces the total circuit area while maintaining the ability to handle special cases in all lanes.
Solution Approach 2:
The shared special case handling circuit serves multiple lanes simultaneously, making it a universal component that handles special case conditions for any lane that detects them. This multi-functional approach allows a single circuit to perform the role of what would otherwise require multiple dedicated circuits, reducing overall system complexity and area.
2Reliability
If special case handling circuits are provided in each lane, then special case conditions are handled correctly, but the overall system complexity and cost increase
Solution Approach 1:
The patent merges the special case handling functionality from multiple lanes into a single shared circuit. The special case handling circuit is provided outside the parallel processing lanes and receives inputs from all lanes, consolidating what would otherwise be duplicate circuits in each lane. This merging reduces the total circuit area while maintaining the ability to handle special cases in all lanes.
Solution Approach 2:
The shared special case handling circuit serves multiple lanes simultaneously, making it a universal component that handles special case conditions for any lane that detects them. This multi-functional approach allows a single circuit to perform the role of what would otherwise require multiple dedicated circuits, reducing overall system complexity and area.
3Area of stationary object
If shared special case handling circuits are used, then circuit area is reduced, but the speed of special case processing may be impacted
Solution Approach 1:
The patent implements preliminary detection of special case conditions in parallel within each lane using detection circuits. While the actual special case processing occurs serially in the shared circuit, the detection phase identifies all special cases simultaneously across lanes. This preliminary parallel action allows the system to prepare for serial processing efficiently, minimizing the impact on overall speed.
Solution Approach 2:
The shared special case handling circuit acts as an intermediary between the parallel detection circuits in each lane and the final processing stages. It receives detection signals from multiple lanes, coordinates the serial processing of special cases, and manages the interaction between parallel detection and serial handling, optimizing the overall processing flow.
Data Source
AI summary
A single instruction multiple data processing pipeline 12 for processing floating point operands includes shared special case handling circuitry 34 for performing any operand dependent special case processing operations. The operand dependent special case processing operations result from special case conditions such as operands that are denormal, an infinity, a not-a-number and a floating point number requiring format conversion. The pipeline 12 may in some embodiments be stalled while the operands requiring special case processing are serially shifted to and from the shared special case handling circuitry 34. In other embodiments the instruction in which the special case condition for an operand arose may be recirculated through the pipeline with permutation circuitry 86, 94 being used to swap the operands between lanes in order to place the operand(s) requiring special case processing operations into the lane containing the shared special case handling circuitry 98.


