Cloud computing resources from Amazon Web Services (AWS) can often be a major business expense, costing companies anywhere from several hundred to hundreds of millions of dollars per year.
AWS offers discount programs and flexible pricing models designed to help companies reduce their spend on cloud resources and services. Choosing the right pricing model is one of the biggest keys to AWS cost management.
To help you make the most of AWS cost optimization, let’s take a detailed look at two pricing models: EC2 Savings Plans and EC2 Reserved Instances.
What are AWS Savings Plans?
AWS Savings Plans are pricing models that offer discounted rates on cloud resources when those resources are committed to for long terms. According to Amazon, AWS Savings Plans can save a business up to 72% on compute usage costs compared to on-demand pricing.
AWS Savings Plans require you to commit to a certain amount of spend on resources over the course of a one-year or three-year term. In exchange for this commitment, Savings Plans offer considerable discounts.
When using Savings Plans to optimize costs on AWS, accurate forecasting is critical; without the ability to predict your resource usage, having to make commitments in advance could mean you end up buying more or less than you actually need.
Types of AWS Savings Plans
There are two main types of Savings Plans for businesses to choose from: Compute Savings Plans and EC2 Instance Savings Plans.
Here’s what you need to know about both plans and how they work:
Compute Savings Plans
Compute Savings Plans provide discounts on a wide range of AWS compute services, including Amazon Elastic Compute Cloud (EC2), AWS Fargate, and AWS Lambda.
This pricing model allows you to apply your savings across various instance families, operating systems, and regions, making it an ideal solution for those that require multiple or changing instance types. In fact, Compute Savings Plans are one of the most flexible discount instruments offered by AWS, with only Convertible Reserved Instances (CRIs) offering more flexibility.
Compute Savings Plans offer three payment options: an All Upfront plan, a Partial Upfront plan, and a No Upfront plan.
The All Upfront plan requires you to pay the entire commitment amount at the beginning of the term. The Partial Upfront plan requires you to pay around half of the commitment amount at the beginning of the term. The No-Upfront plan, on the other hand, doesn’t require upfront payment but also has a higher hourly rate than the other options.
EC2 Instance Savings Plans
The EC2 Instance Savings Plan is a pricing model specifically designed to provide substantial discounts on Amazon EC2 instances. These plans offer savings on a specified instance family and generation within a region.
Purchasing an EC2 Instance Savings Plan requires you to commit to a specific instance family (such as M or C), a specific generation(such as 5 or 6), as well as a specific Region (such as us-east-1, or us-west-2).
Unlike Compute Savings Plans, EC2 Instance Savings Plans don’t provide flexibility across regions, instance types, or instance generations. They also don’t cover other services such as Fargate or Lambda.
What are AWS Reserved Instances?
AWS Reserved Instances (RIs) are another pricing option that provides deep discounts to users who commit to using AWS resources in advance. According to Amazon, using Reserved Instances can save companies up to 75% compared to on-demand pricing.
With AWS Reserved Instances, you have to commit to a specific EC2 instance type (family and generation), operating system, region, and term length.
For Non-Linux OS types, the commitment is also made to a specific instance size. You’ll then have the option to choose from all upfront (AURI), partial upfront (PURI), or no upfront payments (NURI), and the more you pay upfront, the more of a discount you can expect to receive.
Types of Reserved Instances plans
There are two types of AWS Reserved Instances plans, and which offers the most savings will depend on several factors.
Here’s what you need to know about the two types of Reserved Instances plans AWS offers:
Convertible Reserved Instances (CRIs)
Convertible Reserved Instances is the most flexible of the two types of Reserved Instances plans. This flexibility and the potential savings offered by CRIs make them arguably the best discount instrument that AWS offers.
Through the Convertible RI exchange process, you’ll have the option to change the instance attributes within a particular region so long as the resulting exchange is equal or greater to the remaining value of the original CRI(s). This makes Convertible Reserved Instances an ideal solution for businesses with dynamic workloads that may evolve over time.
Standard Reserved Instances (SRI)
Standard Reserved Instances is the original type of Reserved Instances plan, and it is a great choice for any business with steady and predictable workloads.
Standard Reserved Instances require you to commit to either a one-year or three-year term, but you won’t have the option to modify instance attributes other than the availability zone it may be applied against during this term.
In exchange for this commitment, though, Standard Reserved Instances offer the highest savings of the two types. If an SRI is no longer needed upon compute usage type changes, SRIs can be listed on a third-party RI marketplace to offload the remaining term, so long as it is not within the first 30 days or the last 30 days of the commitment term.
Key differences between Savings Plans and Reserved Instances
Savings Plans and Reserved Instances can both offer substantial discounts compared to on-demand pricing, but which option is most cost-effective for your business depends on your specific needs and workload patterns.
Understanding the differences between Savings Plans and Reserved Instances is an important first step to choosing between the two.
Let’s take a look at how these pricing models compare across several key considerations:
Pricing models and payment options
Savings Plans and Reserved Instances both offer similar payment options. With either plan, you’ll have the option to pay all-upfront, partial-upfront, or no-upfront—and in both cases, larger upfront payments lead to higher discounts.
The biggest difference between the two when it comes to pricing is the savings offered by Savings Plans apply to a wider range of AWS compute types. In contrast, Reserved Instances require you to commit to specific instance attributes for EC2 only.
Flexibility and instance coverage
Savings Plans automatically apply discounts to any matching compute usage of the commitment and provide great resource flexibility where no action is taken.
With Reserved Instances, the discount applies to the specific instance attributes you select at the time of purchase, but Linux instances will provide coverage to any matching instance family and generation through Size Flexibility. This means Reserved Instances are ideal for predictable workloads that consistently use the same instance type.
Commitment growth restrictions
Savings Plan growth requires new full-term dollar-per-hour commitments, and coverage cannot be grown on the original expiration date.
Standard Reserved Instances have commitment growth restrictions, making it necessary to purchase additional RIs, either full-term (EC2 Console) or partial-term (RI Marketplace) if your usage exceeds the original commitment.
Convertible Reserved Instances allow for commitment growth through the Convertible RI exchange process on the original expiration date, allowing growth without new, long-term commitments.
Capacity reservation and instance utilization
A Capacity Reservation guarantees that you’ll be able to launch specific instances within a region’s availability zone. Savings Plans do not provide reserved capacity. In other words, with Savings Plans, there’s no guarantee that the resources that you’ve committed to using will always be available exactly when and where you need them.
Reserved Instances, however, can offer capacity reservations, assuring the availability of RIs whenever you need them. The capacity reservation is available when Standard Reserved Instances are purchased at the zonal level only—but not available at the Regional level.
The dedicated capacity provided by Standard RIs makes them a better choice for applications with specific performance requirements. Convertible RIs, however, do not provide capacity reservations.
On-Demand Capacity Reservations can be purchased independently of RIs or Savings Plans if you need to ensure access to specific instances or locations.
Savings Plans and Reserved Instances offer similar discounts based on the different purchase options within each. Standard RIs and EC2 Instance Savings Plans provide similar discounts as they are each very particular to the usage to which they will apply.
Convertible RIs and Compute Savings Plans, due to the flexibility they offer to have changing coverage over the term, offer lower savings than their counterparts.
AWS Savings Plan vs. Reserved Instances: Which one should I use?
AWS Savings Plans cover a wide range of AWS compute services without making you commit to specific instance attributes. This makes them an ideal choice for businesses with dynamic workloads or workloads that require multiple instance types. However, Savings Plans are also immutable, meaning that once you purchase them, you’re stuck with them for the full term.
Savings Plans are easy to work with since they are automatically applied, but they don’t offer nearly as much flexibility as Convertible RIs. For stable workloads with predictable usage patterns, Reserved Instances will also offer the highest savings.
This is why ProsperOps focuses on leveraging Convertible RIs and automatically applies them, so businesses can enjoy the increased flexibility and savings of RIs in a way that is even easier to manage than Savings Plans.
It is important to remember there isn’t a singular solution regarding which of these two pricing models is best for a business. To optimize your AWS spend, you’ll likely need to use a combination of pricing models that are continually adapted and fine-tuned.
What is the best way to optimize AWS costs?
AWS cost optimization is a process that requires constant monitoring and adjustments. The most effective way to optimize costs on AWS is to streamline this complicated and time-consuming process with the power of automation.
With an autonomous discount management tool such as ProsperOps, you can fully automate your AWS cost optimization strategy.
Using machine learning algorithms and advanced data analytics, ProsperOps can continuously analyze your company’s AWS usage patterns to identify inefficiencies and autonomously manage the commitment-based discounts based on the purchase options you select (all upfront, partial upfront, or no upfront).
ProsperOps works around the clock to monitor AWS usage and make cost-saving adjustments in real time, automatically applying discount instruments such as Compute Savings Plans and Convertible RIs. This hands-free approach to AWS cost optimization is able to save your team valuable time while also ensuring this automation continually optimizes your AWS usage for maximum cost savings.
There are several ways to optimize AWS costs, but the easiest and most effective option is to let an AWS cost optimization tool such as ProsperOps do the heavy lifting for you. If you want to reduce your company’s AWS spend without negatively impacting performance or taking time away from your team, ProsperOps may be the answer you’re looking for.
Take control of your AWS spending with Prosper Ops
Choosing between AWS Savings Plans and Reserved Instances is just one of the many crucial decisions that make up a solid AWS cost optimization strategy. It’s also one of the many optimizations ProsperOps can perform automatically.
If you want to take control of your AWS spending and create a cost optimization process that practically runs itself, ProsperOps can help.
Sign up for a free demo of ProsperOps today to see how our industry-leading solution empowers unmatched AWS savings!