Cloud Computing Systems: Easy Business System, Service Model

---

Cloud Computing System 

Cloud computing is a computing paradigm in which computing resources such as servers, storage, networking, and software platforms are delivered over the Internet as services. Instead of purchasing and maintaining expensive hardware infrastructure, organizations can rent computing resources from cloud providers and pay only for what they use. This model significantly reduces the cost of infrastructure management and allows systems to scale dynamically depending on workload demand. Cloud computing has become the backbone of modern digital services, enabling startups, enterprises, and research institutions to deploy applications and process large datasets without building their own data centers.

Cloud platforms provide different service models that allow users to choose the level of control and abstraction they need. 

• IaaS: The most common service model is Infrastructure as a Service (IaaS). In this model, the cloud provider offers virtualized computing infrastructure such as virtual machines, CPUs, RAM, storage, and networking. Users can install operating systems, configure environments, and deploy applications on top of the infrastructure. This model provides maximum flexibility and is commonly used for scientific computing, high-performance computing, and custom server deployments.
• PaaS: The second model is Platform as a Service (PaaS). In PaaS, the provider supplies a development platform that includes operating systems, programming frameworks, runtime environments, and database services. Developers can build, test, and deploy applications without worrying about managing the underlying infrastructure. This model improves developer productivity and simplifies application deployment.
• SaaS: The third model is Software as a Service (SaaS), where complete software applications are delivered over the Internet. Users simply access the application through a web browser or client interface without installing or maintaining the software locally. Examples include cloud-based email systems, collaboration platforms, and enterprise resource planning tools.

Edge Computing 

Egde and FaaS: Modern cloud environments also support new computing paradigms such as Edge Computing and Function as a Service (FaaS). Edge computing moves computation closer to the data source, such as IoT devices, sensors, or local edge servers. This reduces latency, decreases network traffic, and enables real-time processing. On the other hand, FaaS, also known as serverless computing, allows developers to deploy small functions that execute in response to specific events. In this model, the cloud provider automatically manages the servers, scaling, and execution environment, allowing developers to focus only on writing code.

Cloud Renting Model

Cloud computing uses a renting model for resources such as CPU cores, RAM, storage space, GPU accelerators, and network bandwidth. Instead of purchasing these resources permanently, users rent them temporarily based on workload needs. Cloud providers usually offer multiple pricing options. On-demand pricing allows users to allocate resources whenever needed and pay for usage by the hour or second. Reserved pricing allows customers to reserve resources for a longer duration, such as one or three years, often at a significantly reduced price. Regular pricing refers to the standard cost of continuously allocated resources. Spot pricing allows users to bid for unused cloud resources at a lower cost; however, these resources may be reclaimed by the provider when demand increases.

Cloud services operate under Service Level Agreements (SLAs), which define the guarantees provided by the cloud provider. SLAs specify metrics such as system availability, performance, response time, and reliability. For example, a typical SLA may guarantee 99.9% uptime, meaning that the service will be unavailable for only a small amount of time during the year. SLAs may also specify deadlines for recovery, system maintenance policies, and compensation if the provider fails to meet the agreed performance level.

Cloud infrastructure is widely used to host web servers, data servers, and compute servers. Web servers host websites and web applications that need to scale to millions of users. Data servers provide storage and database services for large datasets and enterprise applications. Compute servers support computational workloads such as machine learning, large-scale simulations, scientific modeling, and data analytics. Because cloud platforms allow resources to scale dynamically, they are particularly suitable for workloads with fluctuating demand. As a result, cloud computing has become a fundamental technology supporting modern web services, research computing, and large-scale data-driven applications.

Popular Cloud Service Provider and their cost model

Cloud computing services are provided by several major global companies that offer infrastructure, storage, networking, and computing resources over the internet using a rental-based model. Instead of purchasing expensive servers and storage systems, organizations can rent computing resources such as CPU, RAM, GPU, bandwidth, and storage from cloud service providers. The most widely used cloud providers include Amazon Web Services (AWS), Microsoft Azure, Google Cloud Platform (GCP), Oracle Cloud Infrastructure (OCI), and IBM Cloud. These providers operate large data centers around the world and allow users to access scalable computing resources on demand.

• Amazon Web Services (AWS) is the largest and most widely used cloud platform. It provides compute services through Elastic Compute Cloud (EC2) and storage through services such as Simple Storage Service (S3) and Elastic Block Storage (EBS). The typical cost of AWS storage using S3 standard storage is approximately $0.023 per GB per month. Compute resources such as a virtual machine with about 2 vCPUs and 8 GB RAM cost around $60 to $70 per month depending on the region. GPU-based compute instances used for machine learning or scientific workloads can cost roughly $0.90 to $3 per hour. Data transfer or bandwidth for outgoing internet traffic is typically about $0.09 per GB.
• Microsoft Azure is another major cloud service provider widely used in enterprise environments. Azure provides virtual machines for compute and Azure Blob Storage for scalable object storage. The cost of hot-tier blob storage is approximately $0.018 per GB per month. A typical virtual machine with 2 vCPUs and 8 GB RAM costs around $65 to $75 per month. GPU-enabled instances used for AI or HPC workloads usually range between $1 and $3 per hour depending on the GPU type and configuration. Bandwidth costs for outgoing traffic are around $0.087 per GB.
• Google Cloud Platform (GCP) is also a popular provider, especially for data analytics, machine learning, and large-scale computing applications. Google provides compute services through Compute Engine and storage through Google Cloud Storage. Standard cloud storage costs roughly $0.020 per GB per month. A compute instance with 2 vCPUs and 8 GB RAM costs approximately $50 to $70 per month. GPU instances designed for machine learning or high-performance computing workloads generally cost between $0.80 and $2.50 per hour depending on the GPU model. Network bandwidth charges are usually around $0.10 to $0.12 per GB for outgoing traffic.

All these cloud providers follow a similar pricing model where resources are rented instead of owned. Users can select different pricing options such as on-demand pricing where resources are charged per hour or per second, reserved pricing where users commit to longer usage for lower cost, and spot pricing where unused resources are offered at significant discounts but may be interrupted when demand increases. This flexible pricing structure allows organizations to optimize cost while scaling computing resources based on workload requirements.