Cloud Computing

Cloud computing provides on-demand delivery of IT resources—servers, storage, networking, platforms and applications—over the Internet on a pay-as-you-go basis. By pooling virtualised resources in large data centres, cloud services decouple hardware from software, enabling rapid provisioning, dynamic scaling and centralised management. Key features include self-service interfaces, broad network access from diverse devices, resource pooling under multi-tenant models, rapid elasticity to match workload variations and metered service for transparent billing. Service models range from Infrastructure as a Service (IaaS), where virtual machines and storage are rented, to Platform as a Service (PaaS), which adds managed runtimes and development frameworks, and Software as a Service (SaaS), which delivers complete applications over the web. Deployment options include public, private and hybrid clouds that balance cost, control and compliance. Cloud computing underpins global digital transformation by offering economies of scale, reducing capital outlay and accelerating time to market. It supports a spectrum of use cases—from big-data analytics and machine-learning pipelines to mobile back-ends and enterprise collaboration platforms—while imposing challenges in security, data governance and performance isolation in multi-tenant environments.

Research from Nature Portfolio

No recent Nature Portfolio content available.

Research from all publishers

Recent studies have advanced intelligent and sustainable resource management in cloud environments. A comprehensive review of deep reinforcement learning approaches formulates scheduling as a sequential decision problem, demonstrating that agents can learn to allocate tasks dynamically, yielding lower makespan and higher resource utilisation than classical heuristics. A bio-inspired multi-objective placement method based on the seagull optimisation algorithm jointly minimises energy consumption and network traffic by clustering inter-communicating virtual machines on fewer hosts, achieving up to a 5.5 % reduction in power use and a 70 % cut in network transfers. Complementing these efforts, surveys of software-level energy-efficiency techniques catalogue dynamic voltage scaling, container consolidation and adaptive middleware strategies that collectively curtail data-centre carbon footprints by up to 40 % without compromising quality-of-service guarantees.

Cloud Computing publication trend

The graph below shows the total number of articles in cloud computing across all publications each year (not limited to Nature Index journals).

Technical terms

Virtual machine (VM): A software-emulated computing instance running its own operating system on a physical host, enabling workload isolation and consolidation.

Resource scheduling: The process of assigning computing jobs or virtual machines to available hardware resources to meet performance, cost and energy objectives.

Elasticity: The cloud capability to scale allocated resources automatically up or down in response to demand fluctuations.

Load balancing: The distribution of workloads across multiple resources to optimise utilisation and prevent performance bottlenecks.

Deep reinforcement learning (DRL): A machine-learning paradigm combining reinforcement learning with deep neural networks to make sequential allocation or scheduling decisions in complex environments.

References

1. Cloud Computing Concepts.
2. Deep reinforcement learning-based methods for resource scheduling in cloud computing: a review and future directions. Artificial Intelligence Review (2024).
3. Seagull optimization algorithm based multi-objective VM placement in edge-cloud data centers. Internet of Things and Cyber-Physical Systems (2023).
4. Energy efficiency in cloud computing data centers: a survey on software technologies. Cluster Computing (2022).