> For the complete documentation index, see [llms.txt](https://docs.kinesis.network/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.kinesis.network/understanding-how/how-does-kinesis-network-enhance-customer-value-with-optimized-compute-service.md).

# Plan your compute

Match workload requirements, operational control and cost to the right capacity.

A useful compute decision starts with the workload: what it requires, how long it runs, and how much control you need over its placement.

| Consideration | What to establish                                                                                 |
| ------------- | ------------------------------------------------------------------------------------------------- |
| Hardware      | CPU, RAM, disk, GPU model and VRAM requirements                                                   |
| Runtime       | Startup time, model downloads, sustained load and idle periods                                    |
| Location      | Region preferences and dependencies close to the workload                                         |
| Capacity      | Instance bounds, available servers and shared-grid quotas                                         |
| Operations    | Which organization owns and pays for it; Developers operate apps, admins manage grids and servers |
| Cost          | Serverless usage, dedicated-server charges and external infrastructure costs                      |

## Choose a starting point

**Serverless CPU** is a direct path when you want managed placement for a CPU workload. GPUs currently require a grid, using dedicated, marketplace or connected capacity.

**Dedicated cloud servers** give you explicit server capacity to organize into grids. Their lifecycle is separate from the apps using them.

**Bring your own compute** lets you use existing infrastructure. You remain responsible for the infrastructure's availability and its external costs.

## Measure before expanding

Begin with realistic resource requirements. Inspect logs and usage once the app is running, then adjust the configuration using measured behavior. More instances or different hardware can change both capacity requirements and spend.

For shared grids, check quotas and allocation policy before increasing the app's requirements. A resource being visible does not necessarily mean you can schedule additional workloads onto it.

## Plan the shutdown too

Stop applications that should no longer run. Separately review dedicated machines and outstanding server orders, and release capacity you no longer need. An organization admin can check Cost to understand remaining charges.

[Compare compute options](https://docs.kinesis.network/getting-started/running-llms-with-kinesis-network/understanding-compute-sources.md) · [Understand grids](https://docs.kinesis.network/getting-started/running-llms-with-kinesis-network/understanding-grids.md) · [Get guidance](https://docs.kinesis.network/support.md)


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation by asking a question.

Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters:

```
GET https://docs.kinesis.network/understanding-how/how-does-kinesis-network-enhance-customer-value-with-optimized-compute-service.md?ask=<question>&goal=<user_goal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.
