Cloud ComputingService Models

Spot Instance

Overview

Direct Answer

A spot instance is a cloud computing service that offers access to spare server capacity at discounted rates, typically 70–90% below on-demand pricing, in exchange for the possibility of termination when capacity is reclaimed. Cloud providers allocate these instances from unused infrastructure, making them ideal for workloads that tolerate interruption.

How It Works

Cloud providers maintain a pool of underutilised compute resources and offer them at auction-like pricing that fluctuates based on real-time supply and demand. When an organisation's bid price falls below the current market rate or when capacity is needed for higher-priority workloads, instances are terminated with minimal notice—typically two minutes. Users specify bid prices and interruption tolerance profiles to manage deployment risk.

Why It Matters

Organisations pursuing aggressive cost optimisation can reduce compute expenses substantially whilst maintaining performance for non-critical workloads. This addresses budget constraints in data processing, testing environments, and batch jobs where fault tolerance is inherent to the application architecture.

Common Applications

Typical use cases include machine learning model training, big data analytics processing, batch rendering for media production, and development-stage testing environments. Financial services firms use them for risk simulation, whilst research institutions leverage them for computationally intensive scientific modelling.

Key Considerations

Sudden interruptions require robust retry logic and state persistence mechanisms; unsuitable for stateful applications or time-sensitive transactional workloads. Cost savings must be weighed against engineering complexity required to handle interruption gracefully.

Cross-References(1)

Cloud Computing

More in Cloud Computing