#atom

Distributed computing paradigm that brings computation closer to data sources

Core Idea: Edge computing is a distributed computing model that processes data near its source (the "edge" of the network) rather than in a centralized cloud data center, reducing latency, bandwidth usage, and enabling real-time processing for IoT and mobile applications.

Key Elements

Core Concepts

• Proximity Processing: Computing performed physically close to data sources
• Decentralized Architecture: Distributing workloads across many edge nodes
• Local Data Storage: Temporary or permanent storage at the edge
• Offline Capability: Continuing operation during connection loss
• Intelligence Distribution: Moving AI/ML capabilities to edge devices
• Reduced Backhaul: Minimizing data sent to central data centers

Edge Computing Layers

• Device Edge: Computation directly on IoT devices or sensors
  • Microcontrollers, embedded systems
  • Specialized hardware accelerators
• Local Edge: Small compute nodes near groups of devices
  • Gateway devices, local servers
  • On-premises micro data centers
• Regional Edge: Distributed points of presence
  • Telecom edge facilities
  • Content Delivery Network (CDN) nodes
• Cloud Edge: Cloud provider compute at distributed locations
  • AWS Outposts, Azure Stack Edge
  • Google Distributed Cloud

Key Use Cases

• Industrial IoT: Factory automation and monitoring
• Smart Cities: Traffic management, public safety
• Autonomous Vehicles: Real-time decision making
• Retail Analytics: In-store customer analysis
• Healthcare Monitoring: Patient data processing
• AR/VR Applications: Low-latency rendering
• Telecommunications: 5G mobile edge computing (MEC)

Edge Computing Technologies

• Edge Hardware:
  • IoT gateways
  • Industrial PCs
  • Ruggedized servers
  • GPU/TPU accelerators
• Edge Software:
  • Lightweight Kubernetes (K3s, MicroK8s)
  • Edge-optimized operating systems
  • Containerized applications
  • Edge AI frameworks
• Connectivity:
  • 5G networks
  • Low-power WAN technologies
  • Mesh networking
  • Local connectivity protocols

Edge-Cloud Continuum

• Hybrid Architectures: Workload distribution between edge and cloud
• Cloud Orchestration: Managing edge resources from central cloud
• Data Synchronization: Maintaining consistency across locations
• Progressive Processing: Handling data at appropriate tiers

Challenges and Considerations

• Security: Distributed attack surface
• Management: Device provisioning and updates
• Standardization: Interoperability across platforms
• Resource Constraints: Limited computing power and storage
• Reliability: Operating in harsh environments
• Connectivity: Intermittent network connections

Additional Connections

• Broader Context: Distributed Computing (parent computing model)
• Applications: IoT Architecture (common implementation area)
• See Also: Fog Computing (related intermediate layer concept)

References

1. "Edge Computing: Models, Technologies and Applications" by Pethuru Raj
2. Edge Computing Consortium Standards and Frameworks

#edge-computing #distributed-systems #iot #network-architecture

Connections:

Sources:

• From: Hyper-V - Wikipedia