37
Memory prices climb 500% in 12 months, up to 10x the lowest ever tracked prices — 128GB of DDR5 now $3,399 (lowest-ever price was $329) (www.tomshardware.com)
submitted 6 days ago by to c/gaming@lemmy.world
12 comments fedilink hide all child comments
you are viewing a single comment's thread
view the rest of the comments
[–] 1 point 20 hours ago (1 child)

Seriously, try older models. For a lot of homelab experimentation, you can do fine with 1 to 4 GB ram older pis, like pi 3 or 4. The price is going to be way lower in comparison to a pi 5.

It depends of course on what you're doing, but there's a tendency for people to get the latest and greatest pi when older are just fine and far cheaper.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 2 child comments
    •
  • [–] 1 point 19 hours ago (1 child)

    Yeah I think for the experimentation the older models would be fine and I am looking at getting one for cheaper, there is no stock anywhere at the moment either rn, but I was hoping to have a Jellyfin media server + arr stack and don't think I would be able to run that smoothly with just a Pi. For lighterweight services I could definitely get that, I was even looking at alternative SBCs from aliexpress around same compute power but cheaper.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 2 child comments
    •
  • [–] 1 point 7 hours ago* (last edited 7 hours ago)

    Jellyfin is completely fine on a pi. I believe you just want a 4gb or more pi for that.

    If you do 1080p video at least, I think you can get away with a USB 3 connection to an external hard drive or just a larger micro SD card.

    I just looked up the "arr" stack, lol I get it now. I wouldn't plan on a whole kubernetes style stack unless you treat each pi like its own small vm you'd provision with kubernetes. Just get one pi that fits the minimum requirements for jellyfin, and start there, and then if you really want another part of the stack, you can get a pi that fits the minimum requirements and do that on another.

    it's a difference in methodology. Some people get a big ass rack server and run kubernetes and containers and stuff and spin up tons of little apps on it. With a raspberry pi, you treat one as its own container/vm/whatever. You use whichever has enough resources to do the one task. You let that thing focus on jellyfin or postgres or whatever you want, then if it doesn't have plenty of ram to spare you use another pi for something else.

    it might seem less flexible, but the neat thing is you can usually just swap in and out different micro SD cards between the same pi model if you want to "save" your jellyfin server and experiment with something new on that pi. And also, unlike kubernetes, you have access to GPIO pins and can start wiring your own LEDs or displays or buttons to each and customize them like crazy.

  • permalink
  • fedilink
  • source
  • parent