Main Memory Bus Access for Co-Processors Facing I/O Bottlenecks

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current computer systems face communication bottlenecks due to the lower bandwidths of I/O buses, which restrict the performance of co-processors and I/O devices that operate at faster speeds, necessitating a method to interface these devices through the main memory system for enhanced I/O capabilities.

Innovation Solution

The integration of TeraDIMM technology, which connects co-processors and I/O devices directly to the main memory bus, utilizing a TeraDIMM architecture that includes non-volatile memory and DRAM, allowing these devices to operate at higher bandwidths and bypassing traditional I/O bus limitations, along with a discovery boot loader and driver for initialization and configuration.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Speed

If co-processors and I/O devices interface via the I/O bus, then device compatibility and system simplicity are maintained, but data bandwidth and communication speed are limited

Engineering Contradiction:
Improvedata bandwidthVSAvoidinterface architecture
Core Design Contradiction:
SpeedVSDevice complexity

Solution Approach 1:

The patent applies universality by enabling the main memory bus to serve dual purposes: its traditional function for CPU-memory communication and a new function for interfacing co-processors and I/O devices. This multi-functional use of the high-bandwidth main memory bus eliminates the need for separate I/O buses, thereby increasing data bandwidth for co-processors and I/O devices while maintaining interface simplicity through a unified bus architecture.

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

2Productivity

If co-processors operate at faster speeds, then processing performance is improved, but communication bottlenecks occur due to lower I/O bus bandwidth

Engineering Contradiction:
Improveprocessing performanceVSAvoidcommunication bottleneck
Core Design Contradiction:
ProductivityVSLoss of time

Solution Approach 1:

The patent merges the I/O bus function into the main memory bus system, allowing co-processors and I/O devices to communicate directly through the high-bandwidth main memory bus. This consolidation eliminates the communication bottleneck created by separate low-bandwidth I/O buses, enabling faster processing performance without time loss due to communication constraints.

Inventive Principle:
Principle #5Merging (Combining)

3Productivity

If the main memory bus is used for I/O operations, then I/O performance is enhanced, but bus contention and memory system complexity increase

Engineering Contradiction:
ImproveI/O performanceVSAvoidmemory system architecture
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

The patent applies segmentation by introducing dedicated interface circuitry within the memory controller that segments and manages different types of bus traffic. This interface circuitry handles memory operations and co-processor/I/O device operations separately, allowing the main memory bus to serve multiple functions simultaneously without excessive contention, thereby enhancing I/O performance while controlling memory system complexity.

Inventive Principle:
Principle #1Segmentation

Data Source

PatentEP2673713B1Apparatus and method of interfacing co-processors and input/output devices via a main memory system
Publication Date: 2021.01.27 RAMBUS INC
  • EP2673713B1 patent drawingFigure 1
  • EP2673713B1 patent drawingFigure 2
  • EP2673713B1 patent drawingFigure 3

AI summary

A system for interfacing with a co-processor or input/output device is disclosed. According to one embodiment, the system includes a computer processing unit, a memory module, a memory bus that connects the computer processing unit and the memory module, and a co-processing unit or input/output device, wherein the memory bus also connects the co¬ processing unit or input/output device to the computer processing unit.