SDE Costs#

When you are granted access to the SDE, you will receive a preconfigured environment on Google Cloud . See SDE Structure for details on the standard SDE environment.

As the Google Cloud account owner, you are responsible for managing your usage and associated costs. Cloud services operate on a pay-as-you-go model, meaning you are only billed for what you use. Note that you will be billed for resources that remain active even if you are not using them, such as keeping a virtual machine running around the clock when it’s only being used during business hours.

Cost Sources#

See the Google Cloud documentation for current pricing. The prices included here were current at the time of writing, but they may have changed since.

There are several primary sources of cost for SDE resources:

  • Storage buckets: These are $20/TB/month, billed based on what files are actually stored in the buckets. Buckets can be created and destroyed, but they cannot be turned on and off. As long as a bucket exists, you will be charged for storing the contents.

  • Virtual Machines (VMs): Several projects have preconfigured VMs. VMs have two costs:

    • Time when the VM is “running” (has been started), is billed at a per-hour rate that depends on the VM instance type. The standard SDE environment includes VMs of instance type E2-Standard-8, which cost $0.30/hr.

    • Attached Storage: Storage mounted directly to a VM as a local file system is billed at $0.04/GB/month as long as a VM exists, even when it is “stopped.” The standard SDE environment has 200 GB of local storage for each VM, which costs $8.00/month per VM.

Other resources, such as BigQuery, have pricing models that vary based on the storage and compute resources used. There are also smaller fees related to data transfer, networking, monitoring, and logging services.

Usage Scenarios and Monthly Cost Estimates#

Light Use#

VMs:

Bucket storage:

  • Multiple buckets, 250 GB total stored across buckets

Resource

Rate

Usage Assumptions

Monthly Cost

Virtual Machines (E2-Standard-8)

$0.30/hr

45 VM hours total

$13.50

Persistent VM Disk Storage (3 × 200 GB)

$8/VM

Ingress VM created, used, then deleted

$17

Bucket Storage

$20/TB

0.25 TB

$5

Operational Costs

Fixed

Typical usage

$10

Total Estimated Cost (Light)

$45.50/month

Medium Use#

VMs:

Bucket storage:

  • Multiple buckets, 1 TB total stored across buckets

Resource

Rate

Usage Assumptions

Monthly Cost

Virtual Machines (E2-Standard-8)

$0.30/hr

125 VM hours total

$37.50

Persistent VM Disk Storage (4 × 200 GB)

$8/VM

Ingress VM created, used, then deleted

$25

Bucket Storage

$20/TB

1 TB

$20

Operational Costs

Fixed

Typical usage

$10

Total Estimated Cost (Medium)

$92.50/month

Heavy Use#

VMs:

Bucket storage:

  • Multiple buckets, 1 TB total stored across buckets

Resource

Rate

Usage Assumptions

Monthly Cost

Virtual Machines (E2-Standard-8)

$0.30/hr

770 VM hours total

$231.00

Persistent Disk Storage (6 × 200 GB)

$8/VM

Ingress VM created, used, then deleted

$41

Cloud Bucket Storage

$20/TB

1 TB

$20

Operational Costs

Fixed

Typical usage

$10

Total Estimated Cost (Heavy)

$302.00/month

Monitor VM Use

Leaving your virtual machines (VMs) running when they are not in use can lead to unnecessary costs.
VMs are billed by the hour, even when idle. Always stop your VMs when you’re done working to avoid extra charges.

Prepare Your Purchase Order Information#

While you prepare to request access to the SDE, please have the following information ready. This will help streamline your request and ensure a faster approval process:

  • Blanket Purchase Order (PO): We recommend creating a PO in advance for your project.

  • Justification Text: Include a short description of your research project and how the cloud resources will support it. This is required for PO approval.

  • Estimated Costs: Use the cost calculator above to get an estimate of your expected usage. Include this estimate in your request. You will only be billed for actual usage.

  • Project Details: Add your project name, team members who will access this resource, and any associated funding source.

Having this information ready will allow our team to process your request more efficiently and reduce back-and-forth communications.