Crossbar Interconnect for Discrete Emulation Chips

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current processor-based hardware emulators face inefficiencies in data communication between emulation chips and boards, leading to increased costs and complexity due to the need for additional emulation steps and coordination, which limits the cost-effective increase in processor count within emulation chips.

Innovation Solution

The implementation of a crossbar architecture that allows direct data coupling between multiple discrete emulation chips, enabling parallel processor communication and reducing the need for extensive emulation steps, while maintaining the functionality of larger chips at a lower cost by using smaller, higher-yielding chips.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If the number of processors within an emulation chip is increased to improve computation efficiency, then the computation efficiency is improved, but the manufacturing cost increases due to lower yield from larger chips

Engineering Contradiction:
Improvecomputation efficiencyVSAvoidmanufacturing cost
Core Design Contradiction:
ProductivityVSEase of manufacture

Solution Approach 1:

The patent divides a large emulation chip into multiple smaller discrete emulation chips, each containing a subset of processors. These smaller chips can be manufactured at higher yield and lower cost, while still providing the equivalent computational capacity of a single large chip through parallel operation and coordinated data sharing via cross-chip interconnects.

Inventive Principle:
Principle #1Segmentation

2Adaptability or versatility

If data communication between emulation chips is performed via cable connectors, then communication between boards is enabled, but the communication efficiency decreases due to additional emulation steps and coordination requirements

Engineering Contradiction:
Improvecommunication capabilityVSAvoidcommunication efficiency
Core Design Contradiction:
Adaptability or versatilityVSProductivity

Solution Approach 1:

The patent merges multiple discrete emulation chips into a unified emulation system where the chips operate as a single logical unit. By integrating cross-chip communication pathways and coordinating the chips through shared control mechanisms, the system achieves efficient data sharing that mimics intra-chip communication speeds, eliminating the performance penalty of traditional cable-based inter-board communication.

Inventive Principle:
Principle #5Merging (Combining)

Data Source

PatentUS7725304B1Method and apparatus for coupling data between discrete processor based emulation integrated chips
Publication Date: 2010.05.25 CADENCE DESIGN SYST INC
  • US7725304B1 patent drawing
  • US7725304B1 patent drawing
  • US7725304B1 patent drawing

AI summary

A method and apparatus for coupling data between discrete processor-based emulation chips is described. The apparatus is a processor-based hardware emulation integrated circuits (chips) element comprising a plurality of discrete hardware emulation chips, each emulation chip coupled to another emulation chip by a crossbar for coupling data between the plurality of chips. The method comprises providing data to a crossbar from a first discrete emulation chip, selecting the data from the crossbar using a discrete second emulation chip, and storing the data in a data array in the second discrete emulation chip.