What’s the difference between On-Demand Instances and Reserved Instances?
On-Demand and Reserved Instances are two pricing models for Amazon Elastic Compute Cloud (EC2) instances you can provision in the AWS cloud. Functionally, both types are the same. You can choose from several different compute and memory configurations for different workloads. The difference exists only in pricing. With On-Demand Instances, you pay for computing capacity by the hour or second (with a minimum of 60 seconds) with no long-term commitment. You pay only for what you use, and the instance automatically scales up or down with changing workloads. In contrast, Reserved Instances provide a discounted rate and an optional capacity reservation for your instances. You rent the Reserved Instance for a fixed period at a lower per-second or per-hour rate than the equivalent On-Demand Instance. Spot Instances provide an additional instance pricing model with strict conditions.
Price difference: On-Demand Instance compared with Reserved Instance
On-Demand Instances and Reserved Instances offer different pricing models so that organizations can choose the most appropriate configuration based on their organizational consumption and needs. Configurations of EC2 instances are available with both Reserved and On-Demand pricing.
A Reserved Instance requires a lock-in contract to reserve a full instance for a given time period of either one or three years. On-Demand Instances are billed in hours and seconds and require no lock-in period—they can be turned on or off at will. Compare this This is to renting a car park in the city for a year or on a daily rate. If you work full-time in the city, the yearly rate would be more cost-effective. If you work partly remote, the daily rate may make more sense.
Saving calculations
A Reserved Instance offers cost savings of up to 72% over On-Demand price. In addition, Reserved Instance three-year terms offer much greater savings over one-year terms.
For the same instance type, an On-Demand Instance will always be more costly than Reserved if it is used at capacity for the full time period. For example, if you were to choose a c5d.4xlarge Linux instance with 16 vCPUs, 32 GB memory, and up to 10 Gb network, the On-Demand hourly cost would be 0.96c. Over three years at full capacity, this would equal $25,228.80 in On-Demand pricing. In contrast, the upfront cost for a Reserved Instance for three years is $9,224.28. Even if you only used the instance for half the time instead of its full capacity, you would still save over $3,000.00.
Reserved Instance pricing is not based on use. The price is fixed and locked in. On-Demand Instance pricing is based on use—projected for an equivalent three-year period.
| Utilization |
Reserve demand price for three years |
On-demand price |
Savings |
| 50% |
$9224.28 |
$12614.4 |
$3,390.12 |
| 75% |
$9224.28 |
$18921.6 |
$9,697.32 |
| 100% |
$9224.28 |
$25,228.80 |
$16,004.52 |
How to save more
You can save more by choosing different payment plans and options within Reserved Instances. For example, the EC2 Instance Savings Plan lets you lock in an instance family and region for the term of the plan for a reduced rate. However, if you want the flexibility to switch between instance types during the reserved period, you have to pay slightly more.
Similarly, a full upfront payment offers significant discounts than if you choose to pay monthly fees. For more information about how to purchase Reserved Instances, see Amazon EC2 Reserved Instances Pricing.
Key features: On-Demand compared with Reserved Instance
Both On-Demand and Reserved Instance pricing models offer a lot of flexibility and several options to meet various requirements.
On-Demand Instances
On-Demand Instances offer different pricing levels for data transfer, load balancing, and Amazon Elastic Block Store (Amazon EBS) add-ons. Some instance types are optimized for EBS by default, whereas others offer support for this function. Organizations who like the flexibility of On-Demand Instances but who need guaranteed availability for critical workloads can request an On-Demand Capacity Reservation in a given Availability Zone for any length of time.
Capacity Reservations mitigate the risk of being unable to get On-Demand Capacity in case there are capacity constraints. If you have strict capacity requirements and are running business-critical workloads that require a certain level of long-term or short-term capacity assurance, there is a solution. Create a Capacity Reservation to ensure that you always have access to Amazon EC2 capacity when you need it, for as long as you need it.
Reserved Instances
Reserved Instance attributes such as instance family, instance type, platform, scope, or tenancy may be modified, depending on the requested change. Some AWS Reserved Instances may also be merged. You can also buy scheduled Reserved Instances for recurring time periods, effectively time-sharing your instances for cost savings.
The Amazon EC2 Reserved Instances Marketplace allows customers to buy and sell existing AWS Reserved Instances. The cost to buy a Reserved Instance with a partial term left may be significantly lower than the cost to buy outright. Keep in mind:
- Standard Reserved Instances cannot be exchanged for another Reserved Instance, but can be bought and sold in the marketplace.
- Convertible Reserved Instances can be exchanged for another Reserved Instance with new attributes, but cannot be bought and sold in the marketplace.
Spot Instances
Amazon EC2 Spot Instances are neither Reserved nor On-Demand Instances. Instead, Spot Instances are EC2 instances you can request for increased savings. These Spot Instances will be interrupted if a request for a Reserved or On-Demand Instance exceeds the current Availability Zone capacity. Spot Instances may be preferred over other instances in specific use cases due to their low cost. Spot Instances can be combined with other instance types for further savings.
When to use: On-Demand Instance compared with Reserved Instance
A Reserved Instance is preferred if you use the reserve capacity for the full duration of the one- or three-year term. You may also find that a Reserved Instance offers cost savings over On-Demand Instances, even if you are only using the instance at half capacity. Reserved Instances also offer guaranteed availability for stop-and-start scenarios. In addition, by utilizing and continuously monitoring Convertible Reserved Instances, you can save more money.
On-Demand Instances are preferred when workloads are subject to unpredictable fluctuations, or when large jobs run. These instances are highly-flexible. Organizations may choose On-Demand Instances first to benchmark their workloads over time, before committing to a long-term instance. Billing can be rolled over from On-Demand to Reserved Instances without interruption in service.
AWS also offers calculators to anticipate instance costs, or can provide a detailed quote for more complex configurations.
Summary of differences: On-Demand compared with Reserved Instance
|
|
On-Demand |
Reserved |
| Cost |
Charged per time unit of use |
Fixed for one-year or three-year period |
| Cost variability |
Cost fluctuates with usage |
Cost is fixed irrespective of usage |
| Availability |
Guaranteed only with On-Demand Capacity reservation |
Always guaranteed |
| On-selling |
Optional; Switch off if you no longer require it and no further charges apply |
You can buy and sell Reserved Instances on the AWS Instances Marketplace |
How to get started with EC2 instances?
Amazon Elastic Compute Cloud (Amazon EC2) offers the broadest and deepest compute platform. It has more than 600 instances and your choice of the latest processor, storage, networking, operating system, and purchase model to help best match the needs of your workload.
Amazon EC2 also offers a range of different services for optimizing the cost savings of your EC2 instances. Examine instance type details to ensure your organization is making the most of your cloud compute resources. Combining Reserved, On-Demand, and Spot instances helps drive down costs.
Get started with EC2 instances on AWS by creating an account today.
