SoC Compute Subsystem Isolation for Server Resource Management

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current server infrastructure is costly and inefficient due to the need for additional server systems to provide more computing resources, leading to increased expenses and potential network congestion when multiple clients access shared resources, which can result in performance issues and security risks.

Innovation Solution

Implementing a System on a Chip (SoC) with dedicated server compute and network compute subsystems, each with isolated processing and memory resources, to manage network traffic and provide compute services without significantly increasing costs, thereby reducing jitter and security concerns.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If additional server systems are added to provide more computing resources, then computing capacity is improved, but infrastructure cost increases significantly

Engineering Contradiction:
Improvecomputing capacityVSAvoidinfrastructure cost
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

The patent combines multiple server functionalities into a single SoC device. The SoC integrates a first compute subsystem for host computing services and a second compute subsystem for virtual computing services, along with shared resources including memory, I/O interfaces, and communication interfaces. This consolidation provides multiple computing resources while reducing infrastructure costs and complexity compared to separate server systems.

Inventive Principle:
Principle #5Merging (Combining)

2Productivity

If multiple clients share the same server resources, then resource utilization is improved, but network congestion and performance issues occur

Engineering Contradiction:
Improveresource utilizationVSAvoidnetwork performance
Core Design Contradiction:
ProductivityVSReliability

Solution Approach 1:

The patent segments server resources into isolated compute subsystems within the SoC. The first compute subsystem handles host computing services while the second compute subsystem handles virtual computing services. Each subsystem has dedicated processing resources, preventing network congestion and performance issues by isolating client traffic streams while maintaining high resource utilization through the shared architecture.

Inventive Principle:
Principle #1Segmentation

3Productivity

If computing resources are shared across multiple clients, then cost efficiency is improved, but security risks increase

Engineering Contradiction:
Improvecost efficiencyVSAvoidsecurity risks
Core Design Contradiction:
ProductivityVSObject-affected harmful factors

Solution Approach 1:

The patent implements security isolation through segmented compute subsystems. The first compute subsystem and second compute subsystem are physically separated within the SoC, with each having dedicated processing resources. This segmentation provides security boundaries that prevent unauthorized access between clients while maintaining cost efficiency through shared memory, I/O, and communication resources.

Inventive Principle:
Principle #1Segmentation

4Device complexity

If a single chip is used to provide multiple computing services, then manufacturing cost is reduced, but device complexity increases

Engineering Contradiction:
Improvemanufacturing costVSAvoidcomputing service capability
Core Design Contradiction:
Device complexityVSAdaptability or versatility

Solution Approach 1:

The patent implements a universal SoC architecture that provides multiple computing services through a single device. The SoC includes configurable compute subsystems that can be tailored for different computing needs, along with shared memory, I/O interfaces, and communication interfaces. This multi-functional design reduces manufacturing costs compared to multiple specialized devices while maintaining high adaptability through configurable resource allocation.

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

Data Source

PatentEP3234798B1System on a chip comprising multiple compute sub-systems
Publication Date: 2023.04.05 AMAZON TECH INC
  • EP3234798B1 patent drawingFigure 1
  • EP3234798B1 patent drawingFigure 2
  • EP3234798B1 patent drawingFigure 3A

AI summary

Embodiments can provide additional computing resources at minimal and incremental cost by providing instances of one or more server compute subsystems on a system-on-chip. The system-on-chip can include multiple compute subsystems where each compute subsystem can include dedicated processing and memory resources. The system-on-chip can also include a management compute subsystem that can manage the processing and memory resources for each subsystem.