Round Robin Scheduling Masking Circuit for Priority Arbitration

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

As the number of entries in a scheduling system increases, the complexity of priority arbitration in integrated circuits (ICs) grows, requiring more area and causing longer delays due to the need for additional priority encoders and longer logic paths.

Innovation Solution

A scheduling system that uses a mask generator and priority encoder circuit with a find-first-set bit circuit to determine priority based on previously granted entries, allowing a single priority encoder to handle multiple entries efficiently, reducing area and delay by masking new requests with previously granted entry information.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If priority arbitration is performed for a larger entry count, then access to the shared resource is improved, but area requirements and logic path delay increase

Engineering Contradiction:
Improveaccess to shared resourceVSAvoidarea requirements
Core Design Contradiction:
ProductivityVSArea of stationary object

Solution Approach 1:

The patent segments the priority arbitration process into multiple stages: generating masks from previous grant results, masking original requests to produce masked requests, selecting between original and masked requests, and encoding the selected requests. This segmentation allows the system to handle larger entry counts without proportionally increasing area or delay

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent performs preliminary actions by generating masks from previous grant results before processing new requests. These masks are prepared in advance and used to efficiently filter and select requests, reducing the complexity of priority arbitration for large entry counts

Inventive Principle:
Principle #10Preliminary action

2Productivity

If priority arbitration is performed for a larger entry count, then access to the shared resource is improved, but logic path delay increases

Engineering Contradiction:
Improveaccess to shared resourceVSAvoidlogic path delay
Core Design Contradiction:
ProductivityVSLoss of time

Solution Approach 1:

The arbitration process is divided into separate sequential steps (mask generation, request masking, request selection, priority encoding) that can be implemented with optimized logic paths at each stage, preventing the cumulative delay that would occur in a single large-scale arbitration circuit

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent introduces masks as intermediary elements that mediate between original requests and priority encoding. These masks enable efficient request filtering and selection, reducing the logic path delay compared to direct priority encoding of all requests

Inventive Principle:
Principle #24Intermediary (Mediator)

Data Source

PatentUS10505860B1System and method for round robin scheduling
Publication Date: 2019.12.10 XILINX INC
  • US10505860B1 patent drawing
  • US10505860B1 patent drawing
  • US10505860B1 patent drawing

AI summary

A scheduling system includes a request masking circuit configured to receive a plurality of original requests for priority arbitration among a plurality of entries, the plurality of original requests include a last original request and a first original request following the last original request. A last mask associated with a last grant result for the last original request is received from a mask generator circuit. A first masked request is generated by applying the last mask to the first original request. A request selection circuit is configured to generate a first selected request based on the first original request and the first masked request. The mask generator circuit is configured to generate a first mask based on the first selected request. The first mask is associated with a first grant result for the first original request.