The thing is, most mini PCs with dedicated GPUs seem to come from brands like Minisforum, Geekom, and Beelink, but they add a lot of cost and power consumption. On the other hand, newer iGPUs have gotten a lot better. I'm curious whether the performance trade-off is worth the savings, especially for someone just starting out with self-hosting.

Here's what I'm wondering: Have you found that a dedicated GPU actually makes a difference for home lab workloads, or does integrated graphics handle most virtualization and containerization tasks just fine?

I'd love to hear if anyone has real-world experience or suggestions on this setup, especially if you've tried both. What about power efficiency? does a GPU-equipped mini PC draw a lot more power during idle or light loads? And if you're running local LLMs, what hardware did you settle on and why?

Also, I'm curious if anyone has experience with upgrading RAM and NVMe storage on these compact little pcs. That seems like a smarter investment than GPU power for my use case, but I want to make sure I'm not missing something important. Share your thoughts and help me make the best decision i can...

Download the VMware OVA

Download the VMware OVA from Flatcar:

https://www.flatcar.org/releases/

Choose the VMware/vSphere OVA image.

Create your Butane configuration

Create your normal Butane configuration file (example here):

variant: flatcar
version: 1.0.0

passwd:
  users:
    - name: linuxadmin
      groups:
        - sudo

Save it as:

flatcar.bu

Convert Butane to Ignition

Generate the Ignition JSON:

butane --pretty --strict flatcar.bu > flatcar.ign

Verify the file starts with JSON:

head flatcar.ign

You should see something similar to:

{
  "ignition": {
    "version": "3.3.0"
  }
}

Base64 encode the Ignition file

Encode the generated Ignition file:

base64 -w0 flatcar.ign > flatcar.ign.b64

View the contents:

cat flatcar.ign.b64

Copy the entire output.

Deploy the OVA

Deploy the Flatcar OVA in vSphere.

You can either populate the Ignition fields during deployment or add the values later as Advanced Parameters before the VM is ever powered on. But populating values manually during the OVA install is tedious and doesn't scale very well. You can leave the following screen blank during the deployment:

Configure Ignition using Advanced Parameters

This is the part that you want to pay attention to the details. You need to create the following two advanced configuration parameters:

guestinfo.ignition.config.data
guestinfo.ignition.config.data.encoding 

With the VM powered off:

VM Settings > VM Options > Advanced Parameters

Add:

Both parameters are needed. If you forget to add the base64 advanced setting for the encoding value, you will see errors when it tries to load the config, like this:

Without the encoding parameter, Flatcar will attempt to parse the base64 string as JSON and fail with errors similar to:

error at line 1 col 2: invalid character 'e'

VMware network interface naming

One difference I found out between Proxmox and VMware is interface naming.

In my environment VMware name the NIC as:

ens192

Make sure of the the interface name before building static network configuration.

Example:

Name=ens192


Address=10.1.149.20/24
Gateway=10.1.149.1
DNS=10.1.149.10
Domains=cloud.local

Be careful with Ignition file downloads

My Ignition file contained:

contents:
  source: https://extensions.flatcar.org/extensions/docker-compose.conf

and similar remote downloads.

These worked in my Proxmox deployment but failed during VMware deployment because Ignition attempted to fetch them before networking and DNS were fully available.

The symptom looked like:

lookup extensions.flatcar.org on :53

If you encounter this, deploy without the remote contents.source entries first and move those downloads to a systemd service that runs after:

network-online.target

Verify Ignition

After booting:

journalctl -u ignition-fetch-offline.service -b

or

journalctl -b | grep -i ignition

Successful processing should show:

using OVF environment from guestinfo
config successfully fetched

Hopefully, this forum post will help any who might be looking for information on how to deploy Flatcar in VMware vSphere. It is definitely doable but the process differs quite a bit from Proxmox.