Compute
Elastic Compute Service
Scalable virtual machines with flexible configurations for any workload
Overview
Why Elastic Compute Service?
ECS gives you the broadest choice of compute in the region — from cost-optimised burstable instances to memory-intensive bare-metal servers. Built on our Nitro-equivalent hypervisor, every instance ships with hardware-level isolation, dedicated bandwidth, and sub-second metrics.
Pricing
Straightforward pricing
New accounts receive $200 in free credits applicable to any ConnectVM service.
On-Demand
Pay by the second. No upfront commitment. Ideal for unpredictable workloads.
Reserved
Commit to 1 or 3 years and save up to 72 % compared to On-Demand pricing.
Spot
Bid on unused capacity and cut costs by up to 90 % for fault-tolerant workloads.
Benefits
Built for reliability at scale
Enterprise-grade security
Hardware-enforced isolation on every instance. Instances run in private VPC by default with encrypted storage volumes.
Consistent performance
Dedicated EBS bandwidth and low-latency networking ensure your applications never compete for resources.
Global reach
Deploy instances across 12 regions and 36 availability zones. Place workloads close to your users.
Auto Scaling built-in
Scale compute capacity up or down in response to demand automatically — no manual intervention required.
Features
Everything you need, nothing you don't
Instance families
General-purpose, compute-optimised, memory-optimised, storage-optimised, and GPU instances — all available on demand.
Persistent block storage
Attach high-performance SSD volumes to any instance. Snapshots are automatic and geo-replicated.
Enhanced networking
Up to 100 Gbps network bandwidth with SR-IOV support for the lowest latency between instances.
Confidential computing
Run sensitive workloads inside hardware-based trusted execution environments with memory encryption.
Live migration
Infrastructure maintenance happens transparently. Your instances stay running with zero downtime.
Built-in monitoring
CPU, memory, network, and disk metrics collected every 10 seconds — no agent required.
Use cases
What will you build?
Web & app hosting
Run multi-tier web applications with auto-scaling load-balanced compute behind a managed VPC.
Database workloads
High-memory instances provide the IOPS and RAM needed for OLTP, data warehousing, and in-memory caches.
HPC & batch jobs
Cluster instances with low-latency interconnects for MPI workloads, genomics, and financial modelling.
Dev & test environments
Spin up isolated environments in seconds, run your CI pipeline, then terminate — pay only for what you use.
Resources
Get started
Ready to start?
Create your account and get $200 in free credits.
