Image credit: Unsplash

Building a vSphere 6.5 Homelab 2017

For Under $600, If Possible…

Image credit: Unsplash

Building a vSphere 6.5 Homelab 2017

For Under $600, If Possible…

I’ll be using this homelab to explore vRealize Automation using my VMUG EVALExperience subscription. I’d like this homelab not to take up a ton of power when running 247. Lastly, I’d like the server to look nice if it was out in the open.

Wish list

  • Under $600
  • 4+ core CPU
  • Enough RAM to handle a handful of VMs
  • Very small footprint (no 2U rack servers)

My Build

  • Intel NUC NUC6i7KYK ($325 - ebay)
  • Samsung 850 EVO ($110)
  • Crucial 16GB RAM ($145)
  • TOTAL COST: $580



Intel NUC

The Intel NUC NUC6i7KYK has a i7 6770HQ CPU. This i7 has 4 cores and 8 threads to run a handful of VMs on. All other NUC’s (including other i7s) only have 2 cores.

With a 120 watt PSU, including reports of it utilizing less than 20 watts idle, makes it a perfect server to run 247.

Also, it is SMALL. I think the specs and promo pics don’t do it justice.


This Intel NUC has everything I need.


A Samsung 850 EVO with 250GB is the perfect compliment for my ESXi host. The NUC supports x2 M.2 SSD HDDs and this will be a great start. I also can use the SSD for a local VMFS datastore that is stupid fast. I plan on using this for all of my mission critical VMs.


Using a single stick of Crucial 16GB RAM, I leave the option open to expand to a MAX of 32GB if needed.


This setup will be plenty to get my feet wet with vRealize. It’s also built in a way that I can add more hosts (VMware Datacenter Cluster) or expand my existing server hardware. For now, this will be perfect to leave in my living room without overtaking any other device. I’m thrilled with the bang for the buck on my new lab.

View my upcoming posts for what it looks like to install vSphere 6.5 U1!

Jim Angel
Everything Computers & Cloud

I’m a Cloud Administrator at General Motors. My primary focus is making Kubernetes easier to use. Part of that includes contributing to open source Kubernetes.

comments powered by Disqus