CodeIEB
Theory Notes/🌐 Topic 2: Internet & Communication Technologies/12.2.1
12.2.1Grade 12

Distributed Processing Within a LAN

Where should the 'thinking' happen — on the server, or on the client device? This subtopic covers the spectrum of client types.

Centralised processing
All (or almost all) processing happens on a central server; client devices mainly just display results.
Distributed processing
Processing workload is shared between the server and client devices.
Client typeProcessing done locallyTypical use case
Thin clientVery little — mostly relies on the server for processing and storageCall centres, kiosks — cheap, easy to manage centrally, but useless without network access
Fat clientMost/all processing done locally, server mainly used for shared data/storageTraditional desktop applications — works well offline, but harder to manage/update across many devices
Smart clientA balance — some local processing/caching plus server connectivity for shared data and updatesModern business applications wanting both offline capability and centralised data

Example

A call centre using thin clients means all agent software runs on a central server — if that server goes down, every workstation is unusable, but IT only has to update software in one place.

💡 Exam Tip

Link your choice of client type to specific business needs: cost and central management favour thin clients; reliability without network dependence favours fat clients; a balance of both favours smart clients.