The stuff Hacker News actually recommends.

GPUs and CPUs · NVIDIA GeForce RTX

NVIDIA GeForce RTX 2080

10owners with an opinion
100%of verdicts positive
3/0positive/negative since 2024
0happy after 3+ years

Praised for performance and value.

What people wrote

Likes it
Oh my gosh. Makes me want to wake up my comfyUI setup on a RTX 2080 egpu which still works like a charm today even. I think the last time I touched it was about two years ago! I always found the commercial stuff to be complete rubbish. I was astounded by how I could just spend a couple of hours using open models and open source software to produce better results.
Likes it
> Consumers in the PC segment expect high performance across a wide range of applications It seems there is a persistent rift between what hardware designers think users want and what users are asking for, only recently being addressed by major OEMs with products such as the MacBook Neo. Nobody I personally know wants a faster CPU; they want competent, faster, native feeling software and more power efficient hardware. CPUs have been plenty fast for at least a decade now, for anything the majority of users and even gamers are doing. I'm playing through the new and notoriously demanding LEGO Batman game on an RTX 2080 and an i7-4790k, a CPU from 2013 that even the base M1 easily beats. The game runs fine. Most people I am confident do not need better than the M1.
Neutral
Even 200B for 3k USD is really good, 128 GB of memory for that price is surprising! Right now using a two RTX 2080 setup for a pilot project, where it runs ollama and qwen2.5-coder (14B quantized version), a more serious step up from there on a budget would realistically be an RX 7900 XTX (24 GB, though I've had issues with setting up ROCm) for 1k EUR or RTX A5000 (24 GB) for 2.5k EUR in the current market. Honestly, I don't even need the best performance for what I'm trying to do, even an Arc A770 (16 GB) for 350 EUR would be enough to iterate, except that it's not actually supported by ollama and lots of stuff out there: https://github.com/ollama/ollama/blob/main/docs/gpu.md (I know ollama isn't the only solution, but it sucks when the tools you like aren't available)
Neutral
It really depends on what you're trying to do. This is sorta _the_ guide on GPUs for DL and has a great decision tree https://timdettmers.com/2023/01/30/which-gpu-for-deep-learni... Personally, I'm limited to an RTX 2080 for my personal projects at the moment, and I find the constraint pretty rewarding. It forces me to find alternatives to the huge models, and you'd be surprised what you can eek out when you pour in the time to tweak models. Of course, good data is also pinnacle.
Likes it
One major benefit and the reason why I use the StableDiffusion tools and models is because I can run them at home on my relatively old NVIDIA 2080 GPU with 8GB of VRAM. Costs me nothing (besides electricity). Depends if you value this kind of freedom in life. You can do some things such as colorizing black and white images with the Recolor model. https://huggingface.co/stabilityai/control-lora
Neutral
It's definitely worth a whirl! Some anti-cheats are more effective at catching this than others. I was using this as my method of 'Gaming on Linux' until Proton became a thing. Lots of experience, indeed, though my memory hasn't aged particularly well. I even had SLi working with two RTX2080s! Hacked drivers and EFIGuard to bypass security things Valorant was the one game I couldn't really manage. Perhaps with more determination, but I lost interest rather quickly. Not that into the game and Proton really hurt my VFIO involvement; the timing was unfortunate. There are some rote edits to the libvirt XML I can't recall. Both to get the nvidia driver to work (if applicable, look for 'code 43') , and to hide the VM state for anti-cheats. You'll generally be well served by your distribution of choice with modern kernels and QEMU/libvirt. I don't know Nix well, but from what I gather, you get to pick a lot... so it shouldn't be a problem. Arch is Arch, it'll be fine being so new!
Likes it
Some gamers with high-end cards are likely doing a bit of mining. If the parents or dorms are picking up the electricity bill, that makes sense. And then there are some cards that are used for gaming and to study deep learning, I guess. It is likely a bit more of a mixed use of compute than just pure mining. To give an example, I'm not a student, but I do have a box with 2 x RTX 2080 sitting in my garage, which I use for hobbyist deep learning research. Sometimes, I also need heating in my garage, it gets cold during the winter. It was pretty easy to flash a NiceHash usb stick. So I boot it from it and it becomes a 1.5kW heater which pays for itself. Not particularly efficient or climate friendly, but oh well.