Geometric Synthesis for Integrated Circuit Adder Logic
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing adder trees on integrated circuits, such as FPGAs, inefficiently utilize programmable logic and routing circuitry, leading to high latency and unused logic resources, as they do not fully leverage the available circuitry for performing addition operations.
Innovation Solution
The implementation of geometric synthesis techniques, where counters such as tetrominoes and pentominoes are used to represent bits to be added, allowing adders to be chained together to efficiently utilize programmable logic and reduce routing circuitry, thereby enabling more densely packed adder circuitry and increased logic availability for programming.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If traditional binary adder trees are used, then addition operations can be performed, but programmable logic is not fully utilized and latency increases
Solution Approach 1:
The patent segments the addition operation into multiple parallel partial sum generators, each handling a subset of input vectors. These partial sums are then combined through iterative addition stages. This segmentation allows better utilization of programmable logic resources by distributing the computation across multiple specialized units rather than a single sequential adder tree, thereby improving both productivity and logic utilization.
Solution Approach 2:
The patent introduces a temporal dimension by performing additions in iterative stages rather than purely spatial parallelism. Multiple partial sums are generated simultaneously and then combined through successive addition iterations, effectively using time to resolve resource conflicts. This dimensional approach allows the same hardware resources to be reused across multiple addition stages, improving logic utilization while maintaining high addition throughput.
2Productivity
If adder trees are implemented, then addition can be performed, but routing circuitry is excessively utilized leading to high latency
Solution Approach 1:
The patent merges multiple partial sum generation operations into a unified iterative addition framework. By combining the generation and addition operations into a single coordinated process using shared routing infrastructure, the patent reduces the total routing path length compared to separate adder tree stages. This merging approach decreases routing latency while maintaining high addition speed through efficient resource sharing.
3Productivity
If traditional adder circuitry is used, then addition operations are performed, but the area required is excessive and logic availability for other functions is reduced
Solution Approach 1:
The patent creates a universal addition framework where the same programmable logic resources are used across multiple addition stages and for multiple input vector combinations. The partial sum generators and addition stages share common logic elements and routing infrastructure, making the circuit multi-functional rather than dedicated to a single addition operation. This universality reduces the total area required while maintaining full addition capability.
Data Source
AI summary
A computer-implemented method for programming an integrated circuit includes receiving a program design and determining one or more addition operations based on the program design. The method also includes performing geometric synthesis based on the one or more addition operations by determining a plurality of bits associated with the one or more addition operations and defining a plurality of counters that includes the plurality of bits. Furthermore, the method includes generating instructions configured to cause circuitry configured to perform the one or more addition operations to be implemented on the integrated circuit based on the plurality of counters. The circuitry includes first adder circuitry configured to add a portion of the plurality of bits and produce a carry-out value. The circuitry also includes second adder circuitry configured to determine a sum of a second portion of the plurality of bits and the carry-out value.