From Wikipedia, the free encyclopedia
Linux is a family of open source Unix-like operating systems based on the Linux kernel, an operating system kernel first released on September 17, 1991 by Linus Torvalds. Linux is typically packaged in a Linux distribution (or distro for short).
Distributions include the Linux kernel and supporting system software and libraries, many of which are provided by the GNU Project. Many Linux distributions use the word "Linux" in their name, but the Free Software Foundation uses the name GNU/Linux to emphasize the importance of GNU software, causing some controversy.
Community icon by Alpár-Etele Méder, licensed under CC BY 3.0
Firstly AMD and NVidia both have a range of cards with different power levels. Getting a card with 8G that uses the 75W available from the PCIe slot should be possible.
A card that is setup for ML isn't necessarily going to work well for desktop use. I currently have a NVIDIA card in my workstation setup for ML and it gives 0.2fps on games. I installed the NVIDIA card because the AMD card died and decided to play with ML on it.
I think the best thing to do is to use a separate PC for ML.
play around with local LLMs and image upscaling
FWIW I did that for a bit https://fabien.benetou.fr/Content/SelfHostingArtificialIntelligence and I stopped doing it. I did it mostly from FOMO and that, maybe, truly, it wasn't just hype. Well I stopped. Sure most of those (then state of the art) models are impressive. Yes there are some radical progresses on all fronts, from software to hardware to mathematics underpinning ALL this... and yet, that is ACTUALLY useful in there? IMHO not much.
Once you did try models and confirm that yes indeed it makes "something" then the usefulness is so rare it make the whole endeavor not worth it for me. I would still do it again in retrospect because it helps to learn but... honestly NOT doing it and leaving others to benchmark, review, etc or "just" spending 10 bucks on a commercial model will save you a LOT of time.
So... do what you want but I'd argue gaming remains by far the best usage of a local GPU.
Rx 580, wonderful card, its cheap, 8gb of vram, and runs all the games you need
Just bought at used Radeon RX 7600 8GB on eBay for $185.
What is your current card? The RX 6600 is comfortably within your requirements but has just 8GB of RAM, and the RX 6700 is just at the edge or slightly outside of your budget with 10GB of ram, which you could probably make work if you find a good deal and undervolt it. On Nvidia's side there's the 3050 8GB and 3060 12GB.
RX 480, thinking of upgrading since ROCm support for it ended a while ago and working around it has become very painful
If your big issue is just getting back into mainstream support, you can try looking for a 7600, since RDNA 3 is just one generation removed from the latest. You'll have to undervolt or limit power and you'll be stuck with 8GB on that budget but you'll probably receive software support for slightly longer than the RDNA 2 6600/6700.
RX 7600, comes in 8gb and 16gb variants. GL getting anything that isn't ancient for under $200
You’re gonna have to go a few years back to hit all the goals you have, but a 3060 12gb or even 3070 if you can get one under your price point is gonna have at least 12gb of ram, be near your tdp, maybe over but you can set a power limit, allow you to work with llms and have good linux support.
At that price you're going intel.
in my experience Intel's gpu drivers on linux suck ass. you only get like 60-70% of the card's performance
The newer the line, the better the rocm, but I think the base 9060 is like 8GB 160W $250