PricingPricingDocsDocsBlogBlogBook a demoBook a demoCompanyCompany (opens in a new tab)
Product
DocsAPI ReferencePricingConsole DemoBlogCompany
Start Freev3.4.7
Wave Sizing

Smart Sizing

Recommends a CPU and memory request and limit for each container from its real use, every 10 minutes, and applies them when you choose.

The challenge

You pay for 10 CPU cores and use 1

CAST AI, 2025 Kubernetes Cost Benchmark Report; kubernetes.io, Resource Management for Pods and Containers.

Average CPU utilization across Kubernetes clusters is 10%, and memory utilization is 23% (CAST AI, 2025). Requests are set once, with a large margin, and rarely changed.

The scheduler places a Pod by its request, and the kubelet reserves that amount on the node, whatever the container really uses (kubernetes.io).

A request that is too small has its own cost. A CPU limit below what the container needs is enforced by throttling (kubernetes.io).

Requests far above use, then sized to useWithout WaveWith Waverequest ≫ usesized to use
Without Wave
CPU requests set by hand: two far above use, one throttledCPUrequestuseapiworkerweblimitthrottled
Requests are set high, with a large margin. One workload has a CPU limit below what it needs, so it is throttled.
With Wave
CPU requests and limits recommended per containerCPUrequestuseapiworkerweblimitrecommended
Wave recommends a request and a limit for each container. The analysis runs every 10 minutes.

Bar heights are illustrative. The 10-minute cycle is from the Wave Sizing docs.

How Wave does it

From real use to a new request
Requests set by hand (default)Request stays as written
  1. Engineer
    writes requests in the manifest, once
  2. ↓apply→
    kube-scheduler
    places the Pod by its request
  3. ↓bind→
    kubelet
    reserves the request on the node
  4. ↓usage→
    kubectl top
    shows real use, and the request does not change
Wave Smart SizingNew recommendation every 10 min
  1. Wave Agent
    reads CPU and memory from cAdvisor on every node
  2. ↓metrics→
    Smart Sizing analysis algorithm
    7+ days of CPU, memory, and SLI metrics
  3. ↓every 10 min→
    New request and limit
    per container, within its buffer, min, and max
  4. ↓apply→
    Wave Core
    applies it by Resize Now or Realtime Resizing
Placement and reservation by request: kubernetes.io, Resource Management for Pods and Containers. Wave Agent, 7+ days, the 10-minute cycle, buffer, min, max, and the two apply paths: Wave Sizing docs.
1

Measure every container

The Wave Agent reads CPU and memory from cAdvisor on every node. First recommendations arrive in 24 to 48 hours and are stable after 7 days (Wave Sizing docs).

2

Recommend a request and a limit

The Smart Sizing analysis algorithm uses 7+ days of usage and SLI metrics. For workloads scaled by HPA or Autopilot, it uses a workload performance model instead. A new recommendation comes every 10 minutes.

3

Apply when you choose

Apply by hand with Resize Now, or turn on Realtime Resizing per workload and per container. Restore Original puts the first values back at any time.

Related in Wave Sizing

Realtime Resizing →Smart Sizing Report →Read the docs →
Book a demoStart free
Product
Wave GPUAI Infrastructure Professional ServiceWaveWave AI Agent2026 Q4
Resources
DocsAPI ReferenceBlogConsole DemoPricing
Company
STCLabBook a demoContact
STCLab© 2026 Wave All rights reserved.
Terms of ServicePrivacy PolicyCookies