Skip to main content
Amazon Elastic Compute Cloud (EC2) offers a wide selection of instance types optimized for specific workloads. Selecting the right EC2 instance type ensures you balance compute, memory, storage, and networking to meet performance and cost objectives.

​
EC2 Instance Families

Every EC2 instance family begins with a distinctive letter that highlights its primary use case. Below is a summary of the major instance families:
Evaluate your application’s resource requirements—CPU, RAM, disk I/O, and network bandwidth—to identify the optimal instance family.

​
Key Benefits of Right-Sized Instances

Choosing the right EC2 instance type delivers:
  1. Performance Optimization
    Match vCPU, memory, storage, and networking to workload demands for minimal latency and maximum throughput.
  2. Cost Efficiency
    Right-size your instances to avoid overprovisioning and reduce AWS charges.
  3. Specialized Workloads
    Utilize GPUs, FPGAs, or large RAM pools for AI/ML training, real-time analytics, or high-performance databases.
  4. Scalability & Flexibility
    Resize or switch instance types seamlessly as demand fluctuates.
  5. Architectural Alignment
    Leverage enhanced networking (ENA), EBS-optimized instances, and Nitro system features where needed.
The image lists five benefits: performance optimization, cost optimization, specialized workloads, scalability and flexibility, and architecture optimization. Each benefit is accompanied by an icon and a number.

​
Instance Sizes and Scaling

Within each family, instance sizes typically double CPU and memory resources at each tier. For example, a “medium” instance might offer 1 vCPU and 4 GiB of RAM, scaling up as shown: Actual vCPU, memory, network bandwidth, and EBS throughput vary by family and generation.
Always consult the AWS EC2 Instance Types documentation before provisioning, as configurations and pricing change frequently.

​
Further Reading

Watch Video