Specialized Processing Block Merging Fixed-Point and Floating-Point Arithmetic

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Integrated circuits with specialized processing blocks typically only implement fixed-point arithmetic operations, requiring additional logic outside the block to perform floating-point operations, which is inefficient as applications increasingly demand both fixed-point and floating-point capabilities.

Innovation Solution

Incorporating a specialized processing block with configurable interconnect circuitry that can perform both fixed-point and floating-point arithmetic operations, including multipliers, adders, and rounding circuits, allowing for efficient implementation of floating-point operations within the block.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If floating-point operators are constructed outside the specialized processing block using general-purpose programmable logic, then floating-point operations can be performed, but the device complexity and computational efficiency deteriorate

Engineering Contradiction:
Improvefloating-point operation capabilityVSAvoiddevice complexity
Core Design Contradiction:
Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The patent merges floating-point operators directly into the specialized processing block with fixed-point operators, creating an integrated architecture that performs both fixed-point and floating-point operations within a single unified block, eliminating the need for external general-purpose logic

Inventive Principle:
Principle #5Merging (Combining)

Solution Approach 2:

The specialized processing block is designed with multi-functionality to handle both fixed-point and floating-point operations using shared resources including operators, interconnect circuitry, and memory elements, allowing one block to serve multiple computational purposes

Inventive Principle:
Principle #6Universality (Multi-functionality)

2Adaptability or versatility

If floating-point operators are constructed outside the specialized processing block, then floating-point operations can be performed, but the computational efficiency deteriorates

Engineering Contradiction:
Improvefloating-point operation capabilityVSAvoidcomputational efficiency
Core Design Contradiction:
Adaptability or versatilityVSProductivity

Solution Approach 1:

By merging floating-point operators within the specialized processing block, the patent enables direct integration of floating-point computational resources with fixed-point resources, allowing efficient processing of mixed arithmetic operations without external logic overhead

Inventive Principle:
Principle #5Merging (Combining)

Solution Approach 2:

The floating-point operators are pre-configured and integrated within the specialized processing block structure, with interconnect circuitry and memory elements prepared in advance to handle floating-point operations immediately when needed, eliminating setup delays

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentEP3452895B1Fixed-point and floating-point arithmetic operator circuits in specialized processing blocks
Publication Date: 2021.03.17 ALTERA CORP
  • EP3452895B1 patent drawingFigure 1
  • EP3452895B1 patent drawingFigure 2
  • EP3452895B1 patent drawingFigure 3A~3B

AI summary

The present embodiments relate to circuitry that efficiently performs floating-point arithmetic operations and fixed-point arithmetic operations. Such circuitry may be implemented in specialized processing blocks. If desired, the specialized processing blocks may include configurable interconnect circuitry to support a variety of different use modes. For example, the specialized processing block may efficiently perform a fixed-point or floating-point addition operation or a portion thereof, a fixed-point or floating-point multiplication operation or a portion thereof, a fixed-point or floating-point multiply-add operation or a portion thereof, just to name a few. In some embodiments, two or more specialized processing blocks may be arranged in a cascade chain and perform together more complex operations such as a recursive mode dot product of two vectors of floating-point numbers or a Radix-2 Butterfly circuit, just to name a few.