Basically the only thing that matters for LLM hosting is VRAM capacity
I’ll also add that some frameworks and backends still require CUDA. This is improving but before you go and buy an AMD card, make sure the things you want to run will actually run on it.
For example, bitsandbytes support for non-CUDA backends is still in alpha stage. https://huggingface.co/docs/bitsandbytes/main/en/installation#multi-backend
Borg via Vorta handles the hard parts: encryption, compression, deduplication, and archiving. You can mount backup snapshots like drives, without needing to expand them. It splits archives into small chunks so you can easily upload them to your cloud service of choice.
You can root Graphene if you want to, right?
IIRC, Android has always had native support for keyboards and mice. I remember connecting a bluetooth mouse to my old Nexus 4 running…Android 4, maybe 5? It worked out of the box. Saved my butt when the touch screen broke. :)
Can’t say I’ve tried this in recent years but I think it still works, yeah?
I’ve never tried it myself, but I think you can run full Linux VMs on Pixel phones already. A quick search brings up https://www.xda-developers.com/nestbox-hands-on/
Anyone have experience with this or similar options? Personally I’ve never used anything more advanced than Termux (which is lean and super cool, but not a full-blown VM).
Yeah. It will almost certainly work, but you might not get the same quality as you get on Android. Check the specs of your specific model of earbuds to see which codecs they support. LDAC should work on Debian, but as far as I know Huawei’s proprietary L2HC codec does not work on Linux.
I’m not 100% sure about this since I’ve never used Huawei earbuds myself.
I only have a vague understanding of what iptv is but I guess it’s time to learn. This looks cool.
Seems weird to replace a 3080 but hey, whatever floats your goat.
I switched from Nvidia to AMD recently. As long as you have a recent kernel you should be fine. If you’re running an old/stable distro you might have issues with mesa, especially if you need OpenCL or ROCm. For general use and gaming it worked for me with no fuss.
I forget the exact terminology but I tried putting it into the most permissive mode available. Is still could not work with external hard drives. This was several years ago so I can’t say what might have changed since then, but I did spend some time troubleshooting and at the time that functionality did not work. I’d read that it was possible in the previous version (maybe 18.04?)
Edit: Come to think of it, it might not have been as simple as “couldn’t access external drives”. It might have had something to do with how my disks were mounted and their permissions and mount points. I remember that I hit a wall at some point and further troubleshooting would have required more surgery on my system than I was willing to attempt.
Yeah, this kind of things drove me batty on Ubuntu. So many things were delivered as Snaps when they just don’t work that way. The funniest one to me was Filebot. It’s a media file naming/organizing tool…that doesn’t have disk access. Are you kidding me, Canonical?
Flatpak is easier to work with, but has similar issues. Great for simple things, but I’m always worried that at some point I’m going to need some features that just won’t work, and then it’s going to be a hassle to migrate to a native installation. And it has no CLI support.
And yeah, the bloat is wild. Deduplication on btrfs (or similar) helps but there’s no getting past the bandwidth bloat.
So probably there will be some systems other than Linux that do use Rust
There’s one called Redox that is entirely written in Rust. Still in fairly early stages, though. https://www.redox-os.org/
Have you used Facebook in the last 5 years?
The UX is godawful. More than half my feed is just random crap suggestions and ads.
Installing Linux after Windows should be fine without disconnecting drives.
The reverse is troublesome. Microsoft’s installer is all too happy to shit on your drives, even the ones you’re not using for installation. But Linux installers are much more friendly to dual-booting and all kinds of complex setups.
Same on macOS. Apple has “case-sensitive HFS+” as an option for UNIX compatibility (or at least they used to) but actually running a system on it is a bad idea in general.
Haven’t heard of Hiren’s BootCD in like 15 years. Good to see it’s still around!
Yeah, I had to disconnect all my SATA HDs to stop the Windows installer from shitting all over them.
I’d be worried about Windows updates doing the same thing now, after the the recent glitch that broke bootloaders.
It was bought out and cleaned up a few years ago. It’s legit again now, though I don’t think it’ll ever really recover from that fiasco.
Chromium itself will. Other Chromium-based browser vendors have confirmed that they will maintain v2 support for as long as they can. So perhaps try something like Vivaldi. I haven’t tried PWAs in Vivaldi myself, but it supports them according to the docs.
Debian still supports Pentium IIs. They axed support for the i586 architecture (original Pentium) a few years back, but Debian 12 (current stable, AKA Bookworm) still supports i686 chips like the P2.
Not sure how the rest of the hardware in that Compaq will work.
See: https://www.debian.org/releases/stable/i386/ch02s01.en.html
Yeah, AMD is lagging behind Nvidia in machine learning performance by like a full generation, maybe more. Similar with raytracing.
If you want absolute top-tier performance, then the RTX 4090 is the best consumer card out there, period. Considering the price and power consumption, this is not surprising. It’s hardly fair to compare AMD’s top-end to Nvidia’s top-end when Nvidia’s is over twice the price in the real world.
If your budget for a GPU is <$1600, the 7900 XTX is probably your best bet if you don’t absolutely need CUDA. Any performance advantage Nvidia has goes right out the window if you can’t fit your whole model in VRAM. I’d take a 24GB AMD card over a 16GB Nvidia card any day.
You could also look at an RTX 3090 (which also has 24GB), but then you’d take a big hit to gaming/raster performance and it’d still probably cost you more than a 7900XTX. Not really sure how a 3090 compares to a 7900XTX in Blender. Anyway, that’s probably a more fair comparison if you care about VRAM and price.