Multiple-Memory DSP Architecture for High-Speed Signal Processing

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current digital signal processors (DSPs) face limitations in achieving higher performance and speed while maintaining precision and accuracy, particularly in high-performance computing applications, where they often fall short of providing the necessary processing capabilities.

Innovation Solution

The development of an integrated circuit device featuring an Application-Specific Digital Signal Processor (ASDSP) with multiple memory banks, comprising a control unit, data unit, and function core, which is optimized for specific mathematical expressions and algorithms, allowing for efficient data processing and execution of digital signal processing tasks.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If traditional digital signal processors are used, then basic processing capabilities are provided, but higher performance and speed cannot be achieved

Engineering Contradiction:
Improveprocessing speedVSAvoidprocessor architecture complexity
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

The processor is divided into multiple independent memory banks (first memory bank, second memory bank, third memory bank) that can operate simultaneously. Each memory bank handles specific data streams, allowing parallel access and eliminating memory bottlenecks, thereby achieving higher processing speed without proportionally increasing overall system complexity

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent introduces a multi-dimensional memory architecture where data can be accessed from multiple memory banks simultaneously along different data paths. This dimensional expansion of memory access allows the processor to handle complex algorithms with multiple data streams in parallel, improving productivity while maintaining manageable complexity through structured organization

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

2Speed

If processing speed is increased, then computational bottlenecks are reduced, but maintaining precision and accuracy becomes more difficult

Engineering Contradiction:
Improveprocessing speedVSAvoidcomputational precision
Core Design Contradiction:
SpeedVSMeasurement precision

Solution Approach 1:

Complex computational tasks are segmented and distributed across multiple function cores, each specialized for specific algorithmic operations. This segmentation allows each core to maintain precise computational logic for its designated function while the overall system achieves high speed through parallel execution of multiple segmented tasks

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The control unit acts as an intermediary that coordinates data flow between multiple memory banks and function cores, ensuring that precision-critical operations are executed in the correct sequence and that data integrity is maintained throughout the high-speed processing pipeline

Inventive Principle:
Principle #24Intermediary (Mediator)

3Productivity

If multiple memory banks are integrated, then data access efficiency is improved, but device complexity increases

Engineering Contradiction:
Improvedata access efficiencyVSAvoidmemory bank integration complexity
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

The memory system is segmented into multiple independent banks, each capable of simultaneous access. This segmentation improves data access efficiency by allowing parallel reads/writes from different memory banks, while the modular nature of each bank keeps individual unit complexity manageable

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

Each memory bank is designed with universal interfaces and control logic that can handle various data types and access patterns. This multi-functionality reduces the need for specialized complex control circuits for each bank, thereby improving data access efficiency while limiting the increase in overall device complexity

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

Data Source

PatentUS9111068B2Multiple-memory application-specific digital signal processor
Publication Date: 2015.08.18 HOWARD UNIVERSITY
  • US9111068B2 patent drawing
  • US9111068B2 patent drawing
  • US9111068B2 patent drawing

AI summary

An integrated circuit device is provided comprising a circuit board and one or more digital signal processors implemented thereon. The digital signal processor comprises a data unit comprising a function core configured to perform a specific mathematical expression in order to perform at least a portion of a specific application and an instruction memory storing one or more instructions configured to send commands to the control unit and the data unit to perform the specific application, and a control unit configured to control the flow of data between a plurality of memory banks and the function core for performing the specific application, and the plurality of memory banks coupled to each of the one or more digital signal processors and comprising at least two or more local memory banks integrated onto the circuit board.