Edge computing

Edge computing is processing performed near the devices or places where data is generated. It reduces reliance on distant centralized data centers for timely processing and response.

Cloud computing: Cloud computing delivers computing resources over networks from centralized, shared data centers. Edge computing shifts some workloads away from the centralized data centers typical of cloud services.
Industrial Internet of Things: The Industrial Internet of Things connects industrial equipment and systems to collect and exchange operational data. Factories use edge processing to analyze equipment data without sending every measurement to distant servers.
Centralized computing: Centralized computing performs most processing in a small number of central systems. It is the architectural opposite of distributing processing toward data sources.
Latency (engineering): Latency is the time delay between an input or request and its resulting response. Shorter network paths can make edge processing useful for time-sensitive applications.
Internet of Things security: Internet of Things security protects connected devices, networks, and data from unauthorized access and attack. Widely distributed edge nodes enlarge the number of systems that must be secured and updated.
Fog computing: Fog computing places networking, storage, and computing resources between end devices and centralized cloud infrastructure. It describes an intermediate layer that can coordinate edge devices with cloud services.
Autonomous vehicle: An autonomous vehicle senses its surroundings and performs driving tasks with limited or no human control. Time-critical perception and control must run in or near the vehicle.
High-performance computing: High-performance computing uses powerful systems, often working in parallel, to solve computationally intensive problems. Large workloads requiring concentrated resources may fit centralized supercomputers better than edge nodes.
Network bandwidth: Network bandwidth is the maximum rate at which a communication link can transfer data. Processing data locally can reduce how much information must cross constrained links.
Kubernetes: Kubernetes is an open-source system for deploying, scaling, and managing containerized applications. Its orchestration model is being adapted to manage workloads across edge locations and cloud clusters.

Linked from 11 pages