Free tool

Savings Plans estimator for Kubernetes

Find out how much to commit, what it saves and how far usage can fall before the commitment loses money. Takes a minute.

Your setup

Monthly compute spend at On-Demand prices ($)

Plan type

Compute (flexible)EC2 Instance (fixed family)

Term

1 year3 years

Share that runs 24 by 7 (the steady base)60%

Share already under Savings Plans or Reserved Instances20%

Share you plan to move to Spot15%

How much of the eligible base to commit80%

Scenario: drop in committed usage20%

Estimated savings from the commitment

$1,000per month

$12,000 per year · 5.0% of compute spend

Commitment to purchase: $4.11 per hour
Committed spend (at On-Demand prices): $4,000
Eligible base (steady, uncovered and not going to Spot): $5,000

What if usage drops?

If committed usage drops 20%, the commitment's result becomes $200 per month. Break-even is a 25% drop: beyond that you pay more than you would on On-Demand.

Estimate with illustrative average discounts, not a quote. The real discount depends on instance family, Region and payment option, which is what a technical assessment measures.

Share: LinkedIn

Want to validate this in your environment?

Leave your email and I will get in touch with next steps. No commitment.

Before you buy

How this estimate is calculated

It is not a measurement of your environment: it is a simple model with average discounts and conservative assumptions that I use as a starting point in assessments.

1.Eligible base. The steady share minus what is already covered and what goes to Spot, because Savings Plans do not apply to Spot instances. Anything that runs only part of the time stays out.
2.Commitment. Applies the chosen coverage to the eligible base. The plan is bought in dollars per hour, at the already discounted price, which gives the "per hour" figure shown (using 730 hours a month).
3.Discount. Uses 25% off On-Demand for Compute (flexible) over 1 year, no upfront. AWS advertises higher maximums, up to 66% (Compute) and 72% (EC2 Instance), but they vary by family, Region and term.
4.Drop risk. The commitment is paid in full even if usage falls. Each point of usage lost costs a full point of On-Demand spend, so break-even happens when the drop equals the discount.

What the estimator does not consider

Upfront payment, negotiated discounts, variation by family and Region, seasonality, Reserved Instance resale and the cost of the team's time. That is why the result is an order of magnitude, and a technical assessment measures the real number.

Want to apply this to your environment?

Book a free 30-minute call. I will look at your setup and tell you where to start.

Book a call