I meant on your own system, running the backup software, since it sounds like Duplicati isn’t the best choice. Borg is a great backup program but it’s made for Linux. Kopia is cross-platform.
Sorry, I don’t have a go-to recommendation for the receiving end. In theory anything that can host WebDAV, SSH/SFTP, or an S3 endpoint would do the job.
What platforms are you working with? Borg might serve you better than Duplicati, but last I checked there’s no official Windows version.
There’s also Kopia. I haven’t used it much personally but I’ve heard good things. It supports a range of upload protocols: https://kopia.io/docs/features/#save-snapshots-to-cloud-network-or-local-storage
Most things should work just fine. There’s not all that much that requires x86.
So the real question is, what does your mom need/want? NextCloud? Jellyfin media server? Ente photos?
Maybe just a media center? Basically an Apple TV on crack?
That’s a good question and it’s hard to say for sure. There’ve been some advancements this year in offloading, using both VRAM and system RAM together. With MoE models and predictive loading you can get pretty good results that way…sometimes. I’ve never tried it myself, though I really should because my 24GB GPU is a serious bottleneck.
You could also think about getting a 512GB Mac Studio (coming soon). Not as fast as a GPU, but maybe faster than a GPU/DDR5 mix? And definitely cheaper per GB.
I realize now that the truly biggest models need way more than $40k of hardware. The biggest Qwen3.8 is 2.4 TRILLION parameters. Yikes.
I don’t think you should go bigger because bigger = slower. It might already be too slow to use. Give a smaller model a try first and see how it runs.
It’ll be slow but it’ll run.
Though even 128GB isn’t enough for the biggest models. 512 would open up more options and the very biggest need like 2-4TB, totally out of scope for personal use.
You could try a medium-big MoE model (Mixture of Experts) with less than 100b total parameters. Maybe start with something modest like Qwen3.6-35B-A3B at 8bit quants. That’ll leave you with plenty of memory left over for context size.
If you’re counting models that can be run locally but are too big to run on consumer hardware, then it’s pretty small. Like, they’re only lagging by a few months. But you you’d need like $40k of hardware to run them effectively.
If you mean just the models that you can run on consumer hardware, like, say qwen3.8-27b, then it’s more like 1-2 years behind. You can compensate for the lack of raw capability with a good harness, agentic loops, and human effort. As much as AI has a reputation for being low-effort, it actually requires a lot of skill to use very effectively.
I just looked this up and apparently Papua New Guinea alone has over 800 languages. https://en.wikipedia.org/wiki/List_of_countries_by_number_of_languages
There are over 7000 languages in the world according to https://www.ethnologue.com/
Well crud. Thanks for the heads-up. First I’ve heard.
Sounds like Quad9 doesn’t actually do ad blocking, so that’s not much of a replacement.
It’s free for a limited number of queries per month. Generally enough for one device (depending on personal usage, obviously). Edit: It’ll still function as DNS after that limit, you just won’t get ad blocking and whatnot.
Mullvad has a free ad-blocking DNS you can use, too. I recommend the “base” version: https://mullvad.net/en/help/dns-over-https-and-dns-over-tls
Rclone does provide a docker image, too: https://rclone.org/install/#docker
I’m probably going to set this up soon myself, as I am in the process of building a little home server for my family.