Cloud computing, the long-held dream of computing as a utility, has the potential to transform the IT industry by making software more attractive as a service and reshaping the design and purchase of IT hardware. Developers can deploy services without large capital outlays or human expenses, avoiding overprovisioning or underprovisioning. Companies can scale resources quickly and cost-effectively, as using 1,000 servers for one hour costs no more than using one server for 1,000 hours. This elasticity of resources without premium pricing is unprecedented in IT. Cloud computing is a popular topic, but confusion remains about its definition and usefulness. The authors aim to clarify terms, provide comparisons, and identify technical and non-technical obstacles and opportunities. Cloud computing refers to both applications delivered as services over the Internet and the hardware and systems software in data centers that provide those services. It includes SaaS and utility computing but excludes small or medium-sized data centers, even if they use virtualization. Three aspects are new in cloud computing: infinite computing resources available on demand, elimination of upfront commitments, and the ability to pay for use on a short-term basis. The key enabler of cloud computing is the construction and operation of large-scale, low-cost data centers, which reduce costs through economies of scale and statistical multiplexing. Cloud computing offers three compelling use cases: varying demand, unknown demand, and batch analytics. The economic benefit is captured by "pay as you go," allowing cost savings even if hourly rates are higher than owning. Elasticity allows matching resources to workload more closely, reducing waste. However, underprovisioning can lead to lost revenue and user dissatisfaction. The authors identify ten obstacles and opportunities for cloud computing. These include business continuity, data lock-in, data confidentiality, data transfer bottlenecks, performance unpredictability, scalable storage, bugs in large-scale systems, scaling quickly, reputation fate sharing, and software licensing. Opportunities include improving architectures, standardizing APIs, enhancing security, reducing data transfer costs, and adapting software licensing models. The conclusion is that cloud computing will grow, and developers should consider it. Computing, storage, and networking must focus on horizontal scalability of virtualized resources. Applications need to scale rapidly and use pay-for-use licensing. Infrastructure software must be aware of running on VMs, and metering and billing must be built in. Hardware systems should be designed at the scale of a container, with energy proportionality and efficient resource management.Cloud computing, the long-held dream of computing as a utility, has the potential to transform the IT industry by making software more attractive as a service and reshaping the design and purchase of IT hardware. Developers can deploy services without large capital outlays or human expenses, avoiding overprovisioning or underprovisioning. Companies can scale resources quickly and cost-effectively, as using 1,000 servers for one hour costs no more than using one server for 1,000 hours. This elasticity of resources without premium pricing is unprecedented in IT. Cloud computing is a popular topic, but confusion remains about its definition and usefulness. The authors aim to clarify terms, provide comparisons, and identify technical and non-technical obstacles and opportunities. Cloud computing refers to both applications delivered as services over the Internet and the hardware and systems software in data centers that provide those services. It includes SaaS and utility computing but excludes small or medium-sized data centers, even if they use virtualization. Three aspects are new in cloud computing: infinite computing resources available on demand, elimination of upfront commitments, and the ability to pay for use on a short-term basis. The key enabler of cloud computing is the construction and operation of large-scale, low-cost data centers, which reduce costs through economies of scale and statistical multiplexing. Cloud computing offers three compelling use cases: varying demand, unknown demand, and batch analytics. The economic benefit is captured by "pay as you go," allowing cost savings even if hourly rates are higher than owning. Elasticity allows matching resources to workload more closely, reducing waste. However, underprovisioning can lead to lost revenue and user dissatisfaction. The authors identify ten obstacles and opportunities for cloud computing. These include business continuity, data lock-in, data confidentiality, data transfer bottlenecks, performance unpredictability, scalable storage, bugs in large-scale systems, scaling quickly, reputation fate sharing, and software licensing. Opportunities include improving architectures, standardizing APIs, enhancing security, reducing data transfer costs, and adapting software licensing models. The conclusion is that cloud computing will grow, and developers should consider it. Computing, storage, and networking must focus on horizontal scalability of virtualized resources. Applications need to scale rapidly and use pay-for-use licensing. Infrastructure software must be aware of running on VMs, and metering and billing must be built in. Hardware systems should be designed at the scale of a container, with energy proportionality and efficient resource management.