November 27, 2023
Compute Commitment And Usage-Based Strategies across GCP
3 min read
Navigating multiple Google Cloud compute pricing models and organizing resources to optimize cost could pose a challenge. Even large corporations have experienced significant increases in cloud costs due to inefficient cost management.
Understanding Compute Commitment and Usage-Based pricing on GCP helps organizations with organized resource allocation, utilization, and cost efficiency.
Understanding Compute Commitment on Google Cloud
With Compute Commitment, an organization agrees to use certain compute resources or a certain spend amount for a period of time in exchange for a discount. Google Cloud offers Committed Use Discounts (CUDs) when businesses purchase committed use contracts (known as commitments).
Businesses can get deeply discounted prices for VM instances in return for 1-year or 3-year commitments. These CUDs can be received in either of the following ways:
Resource-based committed use discounts (or resource-based CUDs)
A specific amount of Compute Engine resources are offered at a discounted price when you commit to pay for those resources for a term of either 1 year or 3 years in a particular region.
There are two types of resource-based commitments.
Hardware commitments: Hardware commitments for resources like vCPUs, memory, GPUs, and local SSDs can be purchased at a discount of up to 70% for memory-optimized machine types and 57% for all other machine types.
Software license commitments: License commitments can be purchased for applicable premium operating system (OS) licenses. It offers discounts of up to 79% for SUSE Linux Enterprise Server (SLES) images, 63% for SLES for SAP images, and 24% for Red Hat Enterprise Linux (RHEL) and RHEL for SAP images.
Resource-based commitments are separate for hardware resources and licenses. You will be billed monthly regardless of usage; and once purchased, commitments can not be canceled.
Compute Engine flexible committed use discounts (or flexible CUDs)
Flexible CUDs provide a greater level of flexibility. You commit to a minimum amount of hourly spend on Compute Engine vCPUs and/or memory for a 1-year or 3-year term. You can use the vCPUs and/or memory in any projects within that Cloud Billing Account, across any region and machine type.
In return for committing to an hourly spend amount, you receive a 28% and 46% discount respectively for 1-year and 3-year commitments. It's important to note that the flexible CUDs are applicable only for memory and vCPU resources, and the over-usage is charged at the on-demand rate that is eligible for SUDs. The commitment fee remains the same even if on-demand prices change.
Resource-based and flexible CUDs can only be applied to VMs that are created using Compute Engine, Google Kubernetes Engine (GKE), Dataproc or Cloud Composer 1.
Compute Engine flexible commitments can only be purchased at a Cloud Billing account level.
Compute Engine flexible commitments are not applicable for GPUs, local SSDs or Spot VMs.
Understanding Usage-Based Pricing on Google Cloud
On Demand Pricing
In Usage-Based or Pay-as-you-go pricing models, businesses only pay for the resources they use. You will not be charged once the resource is terminated. There are no upfront fees and the pricing varies broadly by machine type, zone and usage time.
For example, compute-optimized C2 vCPUs and memory costs $0.03827 / vCPU hour, and C2D vCPUs and memory costs $0.033296 / vCPU hour for the Northern Virginia (us-east4) region.
If you have GPUs attached to your resources, the hourly costs will depend on the GPU model, zone, and the number of GPUs attached. For detailed information, please see Compute Engine Pricing.
Sustained Use Discounts (SUDs)
When you consistently use on-demand resources for more than 25% (one-fourth) of the month you will automatically get sustained use discounts for each extra hour that you continue to use that resource. The discount gets bigger with usage and you can save up to 20% or 30% on the cost of virtual machines that run the entire month depending on the machine type.
In the Google Cloud console for your Cloud Billing account, you can see all the applicable SUDs listed as credits on the monthly cost table report. This credit will offset your monthly resource usage costs. These credits have no cash value and cannot be stored.
SUDs for GPUs are calculated separately from VM machine type, vCPU, and memory discounts. You can receive SUDs only on GPUs of the same model.
Spot VMs are unused compute resources that are sold through a bidding process. These VMs are available at up to 60–91% discounts compared to on-demand prices for machine types and GPUs as well as smaller discounts for local SSDs.
However, Spot VMs can be preempted by Compute Engine if it needs it for other tasks making these VMs suitable only for fault-tolerant workloads.
SUDs are automatically applied to VMs created by Compute Engine and GKE. However does not apply to VMs created using the App Engine (standard and flexible) environments and Dataflow.
SUDs apply to only N1, N2, N2D, C2, M1, and M2 machine types.
Spot VMs cannot be seamlessly transitioned into standard VM without being interrupted and are not covered by any Service Level Agreement.
When developing a cost optimization strategy, businesses should consider the following factors:
Workload requirement: Analyze your workload characteristics before selecting a pricing model.
Resource utilization: The goal while optimizing resources should be maximizing discounts and minimizing wastage.
Cost Monitoring: Regularly monitor your cloud costs and look for areas of improvement.
Amnic delivers a cloud cost observability platform that helps measure and rightsize cloud costs continuously. It is agentless, secure and allows users to get started in five minutes at no cost.
Amnic provides a suite of features such as cost explorer, K8s visibility, custom dashboards, benchmarking, anomaly detection, alerts, K8s optimization and more. Businesses save 25-30% on their cloud costs, even on pre-optimized environments.
Amnic provides precise recommendations across network, storage and compute, based on your cloud infrastructure to identify high costs and industry best practices to reduce them. With Amnic, organizations can successfully build a roadmap towards lean cloud infrastructure and build a culture of cost optimization among their teams.
Visit www.amnic.com to learn more about how you can get started on your cloud cost optimization journey.