Compute services: choosing the right one

The services at a glance

The Compute category of the exam covers eight services. At its center is Amazon EC2, virtual servers that you control; the others either manage EC2 for you or bring compute to places outside an AWS RegionA separate geographic area where AWS runs its services, such as us-east-1 in Northern Virginia, made up of several Availability Zones. Most resources live in one Region; you choose Regions for latency, data residency, and the services you need.Glossary entry. Each service name links to its own tutorial. Containers and serverless compute (Lambda and Fargate) have their own exam categories.

Service	What it does
Amazon EC2	Virtual servers (instances) with your choice of processor, memory, storage, networking, and operating system
Amazon EC2 Auto Scaling	Keeps the right number of healthy EC2 instances running, adding and removing them as demand changes
AWS Batch	Runs batch computing jobs, queuing them and provisioning the compute they need
AWS Elastic Beanstalk	Deploys and runs web applications: you upload code and it creates the instances, load balancer, and scaling
AWS Serverless Application Repository	A catalog for finding, deploying, and sharing ready-made serverless applications
AWS Outposts	AWS infrastructure and services in your own data center, for low LatencyThe time between sending a request and receiving the response, usually measured in milliseconds. Distance and processing add to it; caching and serving users from nearby locations reduce it.Glossary entry or data residency
AWS Wavelength	AWS compute and storage at the edge of telecom providers' 5G networks, for very low latency to mobile devices
VMware Cloud on AWS	VMware environments on AWS infrastructure, for moving VMware workloads without converting them; sold by Broadcom since 2024, with Amazon EVS as the AWS-native option

How to choose

Ask three questions, in this order:

Where must it run? In an AWS Region (most workloads); On-premisesIn data centers or offices that you run yourself, rather than in the cloud. Hybrid architectures connect on-premises systems to AWS, for example through AWS Direct Connect or a VPN.Glossary entry (Outposts); or at a 5G network edge, close to mobile users (Wavelength). Existing VMware environments point to VMware Cloud on AWS.
How much do you want to manage? Full control of servers is EC2; uploading code and letting AWS run the environment is Elastic Beanstalk; deploying an application someone already packaged is the Serverless Application Repository.
What kind of work is it? A web or application tier that must stay available and follow demand is EC2 with Auto Scaling; queued jobs that run to completion, such as simulations or nightly processing, are AWS Batch.

Then choose how to pay: On-Demand for short or unpredictable use, Savings Plans or RI (Reserved Instance)A billing discount for committing to an instance type in a Region or Availability Zone for one or three years. It is not an instance: running instances that match are billed at the lower rate. AWS now recommends Savings Plans instead.Glossary entry for the steady baseline, and Spot Instances for interruptible work.

Reverse mindmap: from the question to the service

Compute: which service?

Servers you control
- Specific OS, instance type, or software on a server leads to Amazon EC2
- Keep a fleet healthy and follow demand leads to EC2 Auto Scaling
- Lowest latency between nodes for HPC leads to Cluster placement group
- Server-bound licenses or dedicated hardware leads to Dedicated Hosts
Pay less for EC2
- Interruptible, flexible workloads leads to Spot Instances
- Steady baseline for 1 or 3 years leads to Savings Plans
- Guaranteed capacity in one Availability Zone leads to On-Demand Capacity Reservation
Let AWS manage more
- Upload web app code, AWS runs the environment leads to Elastic Beanstalk
- Queued jobs that run to completion leads to AWS Batch
- Deploy a ready-made serverless application leads to Serverless Application Repository
Run outside an AWS Region
- AWS services in your own data center leads to AWS Outposts
- Lowest latency to mobile devices on 5G networks leads to AWS Wavelength
- Move VMware workloads without changing them leads to VMware Cloud on AWS

Commonly confused pairs

If the choice is between	Choose the first when	Choose the second when
EC2 or Elastic Beanstalk	You need full control of the servers	You want to deploy code without managing the infrastructure
AWS Batch or EC2 Auto Scaling	Work arrives as jobs that run to completion	A service runs continuously and its load changes
Spot or On-Demand	The work can be interrupted and restarted	The work must not be interrupted
Outposts or Wavelength	Compute must run in your own data center	Compute must be close to mobile users on a 5G network
Dedicated Hosts or Dedicated Instances	You need visibility of sockets and cores for licensing, or host affinity	You only need hardware that no other customer uses

Key takeaways

EC2 is the foundation: choose the instance type for the workload and the purchase option for the usage pattern.
Auto Scaling keeps EC2 fleets healthy and sized to demand, across AZ (Availability Zone)An isolated location within an AWS Region, made up of one or more data centers with their own power, cooling, and networking. The AZs in a Region are linked by fast, low-latency connections, so spreading resources across them protects against the failure of one site.Glossary entry.
Elastic Beanstalk and AWS Batch manage EC2 for you, for web applications and batch jobs respectively.
Outposts brings AWS on premises; Wavelength brings it to the 5G edge; VMware Cloud on AWS runs VMware workloads unchanged.

Compute services: choosing the right one

The services at a glance

How to choose

Reverse mindmap: from the question to the service

Commonly confused pairs

Key takeaways

Sources

Check your understanding

Add LetMeMaster to your Home Screen