Amazon EC2

Amazon EC2 is a web service that provides compute capacity in the AWS cloud. You have several choices of instance types, operating systems, and software packages available. You will need to choose an appropriate mix of instance types based on your specific use cases. These instance types include general purpose, accelerated computing, compute-optimized, memory-optimized,  and storage-optimized instances. For example, you would choose compute-optimized instances for your compute-intensive workloads and accelerated computing instances for GPU-based processing, typically, used for deep learning applications. In addition, each instance type includes one or more instance sizes to match the scalability requirements of your specific workloads.

Amazon EC2 allows you to configure the memory, CPU, instance storage, and your choice of operating system and applications. You can bundle the operating system, application software, and associated configuration settings into an Amazon Machine Image (AMI). Then, use it to provision or decommission multiple virtualized instances using web service calls. EC2 instances can be resized and the number of instances scaled, up or down, to match your requirements or demand. 

Each AWS region comprises several Availability Zones (AZs) at distinct locations, connected by low latency networks. EC2 instances can be launched in one or more geographical locations or regions and also within one or more AZs belonging to a specific region. 

Amazon Elastic Block Storage (EBS) volumes provide network-attached persistent storage to the EC2 instances. Elastic IP addresses allow you to allocate a static IP address and programmatically assign it to an instance. You can enable monitoring on EC2 instances using Amazon CloudWatch, and auto-scaling groups can be created, using the auto scaling feature, to automatically scale available capacity (based on the CloudWatch metrics).