> For the complete documentation index, see [llms.txt](https://docs.lumia.org/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.lumia.org/run-a-node/hardware-requirements.md).

# Hardware requirements

**Who this is for:** operators sizing a host before deploying a Lumia RPC node.

**Every figure below is a working estimate, not a published spec. Confirm before you provision anything at scale.**

### RPC node

| Resource | Minimum                                         | Recommended                               | Notes                                                                                                                                |
| -------- | ----------------------------------------------- | ----------------------------------------- | ------------------------------------------------------------------------------------------------------------------------------------ |
| CPU      | 4 cores, x86\_64                                | 8+ cores                                  | Single-thread performance matters more than core count for sync speed                                                                |
| RAM      | 32 GiB                                          | 64 GiB                                    | Stated as a hard minimum for the execution client, not a soft recommendation                                                         |
| Disk     | 1 TB NVMe SSD                                   | 2×1 TB NVMe                               | Local NVMe only: no network-attached or spinning storage; current chain state is far smaller than 1 TB, but the figure covers growth |
| Network  | 1 Gbps, public IPv4 if serving external clients | 1 Gbps+, meaningful monthly egress budget | Egress scales directly with RPC traffic served                                                                                       |

This is the node type most operators actually want. See [RPC node overview](https://app.gitbook.com/s/chmhPiVNAafGw48c4oU2/run-a-node/rpc-node/rpc-node-overview).

### Before you provision

* Confirm free disk headroom before starting a snapshot bootstrap: running out of space mid-extract is a common cause of a corrupted data directory. See [Node troubleshooting](https://app.gitbook.com/s/chmhPiVNAafGw48c4oU2/run-a-node/node-troubleshooting).