docs/selecting-hardware.md

Rama main · réplica de lectura del repositorio original.

---
title: Selecting hardware
noIndex: false
noContent: false
---

### Hardware options

Every project in Deepnote runs on a machine in the cloud. This means you don't need to worry about any kind of setup. Note that all hardware options have an enabled connection to the internet and therefore keep running even if you close your browser (also see [long-running jobs](/docs/long-running-jobs)).

<Callout status="info">

At this moment Deepnote does not support connecting to your own cloud-based resource (e.g. AWS, GCP), but this is coming soon.

</Callout>

To choose a different machine for your project and monitor its performance, click the selector under **Machine** in the right sidebar. You can then select another machine. More powerful machines are available on paid plans. See [Pricing](/docs/pricing) or [contact us](mailto:help@deepnote.com) for details.

<VideoLoop src="../assets/docs/1tOtitPQRwG3IlOoziUB.mp4" />

# GPUs

Deepnote's GPUs are designed for machine learning workloads. You can change the machine at the notebook level. The current machine selector includes two GPU options:

- GPU (T4): 8 vCPU, 32 GB memory, 16 GB VRAM
- GPU (L4): 16 vCPU, 64 GB memory, 24 GB VRAM

Please [contact us](mailto:help@deepnote.com) if you require more powerful GPUs.