StorageClass Performance Calculator
Compare IOPS and throughput characteristics across Kubernetes StorageClasses.
About StorageClass Performance Calculator
Kubernetes storage classes map to underlying cloud block storage tiers with very different IOPS ceilings and cost structures: gp3 offers up to 3,000 baseline IOPS at low cost, io2 scales up to 64,000 IOPS for demanding databases at a premium, and standard (magnetic/HDD-backed) tops out around 100 IOPS for cold, infrequently accessed data. This calculator filters out classes that can't meet your required IOPS, then recommends the cheapest one that can.
Formula
monthlyCost = volumeSize × costPerGiB
- V
- Volume size in GiB
- c
- Cost per GiB per month for the storage class
- C
- Monthly cost for the volume
Worked example
With the default values below, StorageClass Performance Calculator returns Recommended Storage Class of io2. Change any input above and every figure here is recalculated in your browser.
Inputs used
- Volume Sizeaccepts 1–1,000,000 GiB
- 100 GiB
- Required IOPSaccepts 0–500,000 IOPS
- 5,000 IOPS
- gp3 Costaccepts 0–100 $/GiB/mo
- 0.08 $/GiB/mo
- io2 Costaccepts 0–100 $/GiB/mo
- 0.125 $/GiB/mo
- standard (HDD) Costaccepts 0–100 $/GiB/mo
- 0.045 $/GiB/mo
Results
- Recommended Storage Class
- io2
Step by step
- 1Required IOPS= 5000
- 2Viable storage classes= io2
Assumptions and limits
- The result is derived only from the 5 inputs listed above. Anything not among them is not part of the calculation.
- In particular, this calculator does not take inflation, tax or compounding frequency into account.
How these calculators are built and tested is described on the Methodology page. If a figure looks wrong, please report it.
Frequently Asked Questions
When does gp3 stop being sufficient?
gp3 provides a 3,000 IOPS baseline (extendable up to 16,000 with additional provisioned IOPS on AWS) — workloads sustained above that, like high-throughput OLTP databases, typically need io2 or io2 Block Express.
Is io2 always better since it has higher IOPS?
No — io2's per-GiB cost is meaningfully higher, and most workloads don't need anywhere near 64,000 IOPS. Overprovisioning IOPS you don't use is pure cost waste; size to your actual measured requirement plus headroom.
What workloads still make sense on 'standard' storage?
Cold storage, backups, infrequently-accessed logs, or any workload dominated by sequential throughput rather than random IOPS — its low IOPS ceiling makes it unsuitable for databases or anything with random read/write patterns.