# Storage

* Direct Attached Storage (DAS)
* Network Attached Storage (NAS)

<figure><img src="/files/BxyzLSXuhSHbvvva1c20" alt="" width="451"><figcaption></figcaption></figure>

## Direct Attached

Low latency ephemeral storage, ideal for caching training datasets, model checkpoints and intermediate results. Denvr hosts include up to 30TB with read throughput up to 30 GB/s. On Denvr virtual machines these disks can be formatted and mounted with the following:

```shell-session
sudo mkfs -t ext4 /dev/vdc
mkdir -p /home/ubuntu/direct-attached
mount /dev/vdc /home/ubuntu/direct-attached
```

#### Tips

* Cache your models and datasets here to avoid bottlenecks during training or inference tasks.
* Backup results or checkpoints to our NAS for reuse on other Denvr hosts.

## Network Attached

Reliable distributed network storage for managing VM root disks, saving model checkpoints, and sharing datasets across hosts and even teams. Our network storage comes in **general** and **high-performance** tiers.

### General Storage

Low cost redundant storage for home directories, system files, reference datasets or docker images. Petabyte-scale with read throughput up to 1GB/s.

#### Tip

* Ideal for virtual machine root disks (`/dev/vda`)  and custom images.

### High-Performance Storage

High performance and redundant storage for sharing large model checkpoints and datasets across a cluster of machines or between users. Petabyte-scale capacity with read throughput up to 10GB/s.

#### Tip

* Backup DAS cached model checkpoints or training results here for future analysis or retraining.
* Personal shares are user specific stores that can be accessed from our virtual machine or application services. Default location is `~/personal`.
* Tenant shared can be accessed by any user in an account and is also available from either virtual machines or applications. Default location is `~/tenant-shared`.


---

# Agent Instructions: Querying This Documentation

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

Perform an HTTP GET request on the current page URL with the `ask` query parameter:

```
GET https://docs.denvrdata.com/docs/platform/storage.md?ask=<question>
```

The question should be specific, self-contained, and written in natural language.
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.