304
I Put a Datacenter GPU in My Gaming PC for £200
(blog.tymscar.com)
you are viewing a single comment's thread
view the rest of the comments
view the rest of the comments
this post was submitted on 19 Jul 2026
304 points (100.0% liked)
Technology
86636 readers
2864 users here now
This is a most excellent place for technology news and articles.
Our Rules
- Follow the lemmy.world rules.
- Only tech related news or articles.
- Be excellent to each other!
- Mod approved content bots can post up to 10 articles per day.
- Threads asking for personal tech support may be deleted.
- Politics threads may be removed.
- No memes allowed as posts, OK to post as comments.
- Only approved bots from the list below, this includes using AI responses and summaries. To ask if your bot can be added please contact a mod.
- Check for duplicates before posting, duplicates may be removed
- Accounts 7 days and younger will have their posts automatically removed.
Approved Bots
founded 3 years ago
MODERATORS
Yes.
Lemmy is just by default super AI-hating. I wonder what they felt when Linus Torvalds said that LLM are actually useful tools.
We felt like he's right about it being useful tools but the ethical aspects are hard to ignore. Yes, many ethical aspects.
You're likely conflating the tools with the implementation or the implementers. An open source model on local hardware has next to zero ethical concerns.
There are MANY ethical concerns about the actions of the big players in AI, their data centers, and to some extent, the original, bootstrapped training data sourcing.
Well, yes and no. Even open source models had to consume a lot of power for the training, including the use of non-owned data
Many, many, many things "consume a lot of power" which is both common and benign. A whirlpool tub "consumes a lot of power" for example. So does an electric stove. And the power can be from any source, such as solar, hydro, or wind.
That's why I specified there's minor issues that could be argued about the original data sources. However, that bird has flown the coop. There's no putting that genie back in the bottle, and no way to undo it. You could argue the company's responsible owe every single person on the planet some form of restitution, which I do, but I don't think that qualifies as an ethical concern since it's universal. No one person was harmed more than another, it was all publicly available data - so the public should benefit from the result. Refusing to use it only hampers yourself, at no detriment to those perceived as wrongdoers.
I rarely see people agonize over the ethics of playing a game on high settings for hours because of the frivolous use of electricity.
Right? Yet using AI takes less electricity than that. The concerns are being wildly exaggerated, overblown, and I suspect it's being done intentionally as a psychological attack on societies to hinder competing nations from faster adoption.
I don't think so, no?
Okay. So, imagine I'm an indie game developer. I've got no artistic talent, no funds, and am just making a game that I want to play, not one I think will make any money.
With me so far? I download a free, open source, open weights model on my laptop. I install the open source tools to allow the LLM to read my project files, and I give it a thorough description of the game I'm designing. I work on some aspect in the foreground, while in the background my GPU is utilized to achieve some task I've assigned the AI. When it's done, I review the work, commit the changes, and assign it a new task.
What are the ethical aspects you can't ignore in my use case?
If you don't intend to sell or distribute that game, the only aspects left that I can think of are:
Even if I did intend to sell the game, what change is there? Everything created was something that did not exist prior my prompt to create it.
I'm making myself dumber in the same sense that learning to read & write weakens my memory skills, using a calculator weakens my math skills, or driving a car weakens my physical stamina. Everything is a trade off, and like in everything else, balance is the key consideration.
Correct, I personally have a solar array with a battery system. I typically have an overage, and feed my power back into the network. Even if I didn't, though, using my GPU for AI uses less electricity than playing a game with it, and even that is dwarfed by microwaving food, using a hair dryer, clothes dryer, etc. Power consumption is a common concern that is shared with everything, so there's no justification to make it an ethical one just for AI.
In my case, I'm using Qwen, and Kimi, which are both models trained by other models - essentially compressing the original training so a good analogy, if these were people, is my models were trained by professors who themselves trained themselves by reviewing the entirety of the Internet. It's important, though, to highlight the fact that as neural networks this training is not retention of data - it's simply adjusting the weights of simulated synapses based on the tagging of input.
So many people seem to think that training means the model has like a compressed database of source material, but that is not the case - even on the original models that trained using pirated data, at no point is the training data incorporated into the model. It informs the model, in the same way you are informed about art by watching a Disney movie. Sure, you might be able to recreate something you saw from memory if you focused on doing that, but it's not like you can make a perfect copy from your memory. The LLM are NOT idetic.
Power consumption for AI is way higher than for games…your gpu is rarely at full usage when running a game, only really if you’re one of the few people that tries to get every single bit out of your game with the highest level graphics and even then you’re not using the full amount because the game would stutter. It doesn’t matter if an llm stutters. They use the full power of your gpu no matter what you’re doing.
And the models being trained on other models mean that local LLMs use more energy than a model like Claude. You have all the energy of multiple data centers training one model and then you go and train a model off of that.
You are mistaken. Power consumption is lower because the VRAM is what's primarily being utilized, not the GPU. Not only that, but GPU & VRAM are used non-stop while gaming, but only intermittently while performing LLM inference. And it's asynchronous, only processing when replying, while the rest of the time it's idle, while waiting for the next prompt. Gaming uses a LOT more electricity than local LLM usage.
If you're going to count the processes that had to occur prior to the current usage, then gaming uses a magnitude more electricity because of all the electricity of the developers who made the game, and the artists, and the sound designers, and the composers, and the compilers, etc.
Let's just leave it at both use substantial energy, and both are difficult to justify. I still personally love and will use both, as a game developer with a focus on novel AI usage in games.
The temperature my computer room gets to when gaming vs using a local model says otherwise. I’ll gladly plug in a measuring tool and show you the data if you want.
That depends entirely on the hardware, the game, the model, and the prompt. I'm sure you could find a combination where the LLM power usage is higher than the game, just as I could do the opposite. The point is, they both use power, but only one of them is disparaged.
The one you’re using with power in your own house already used enough power to run a city, so yeah…of course that one is getting disparaged.
And no, it doesn’t depend on the hardware. Thinking that a local llm only uses a portion of your gpu while games max them out is just ignorant. Games do not have to load up the entire vram and run the gpu at full power just to run a basic visual novel. An llm does both of those things, anytime you are using it. It’s absolutely nothing like a game.
Yeah, and the games you're playing spent literally years in development, with hundreds, if not thousands of developers, using their computers to create it. Not to mention rendering CGI for FMV, nor the servers to host multiplayer.
It definitely does depend on the hardware. You can use a local LLM without even a GPU, and using only a CPU. You've obviously never played Doom: Eternal, or Final Fantasy VII, or Grand Theft Auto V. What kind of lame person is paying a visual novel on a gaming PC? Talk about ignorant.
I'm just going to block you now. I'm tired of your ignorance claiming false things. Have a nice, AI-free life while playing your visual novels.
The visual novel was a joke reference to an LLM using up an entire gpu. It’s quite sad that you’re so addicted to LLMs that you don’t even notice your biases.
All of those things you listed take a fraction of the power of these models.
You keep making your case around a veeeeeeeeeryyyyy specific type of user. It's hard not to think it's actually a specific person you're talking about lol. Hardly a common scenario. 😆
Yes, it's me! I'm the very specific type of user, and I'm being swept up in the discrimination in the mad rush to hate everything AI.
You say you are being swept up like some sort of unrelated bystander but it doesn't look like your justification is too different from the typical ai user's justification.
Justification of what? Using the incredible technology created via the shared culmination of all human knowledge, on personally owned hardware, using clean, renewable energy? What needs to be justified in that?
A good amount without their consent yeah. You tried to justify it earlier in the comment chain by saying it's not an exact reproduction and that the cat's already out of the bag, which is what ai bros typically say to cope with that aspect.
You don't need consent to learn from something, yeah? Or do artists need consent to learn by watching their movies? Do authors need consent to learn by reading books? It's machine LEARNING, not machine copying.
Yes, courses and materials coming with terms and conditions for its use is not really a new concept for most reasonable people.
Then, I guess if a course or material owner agrees with your logic, and have a complaint then they should file a lawsuit and let the legal process carry out. In my own judgement, nothing unethical occurred and complaints such as your own are nonsensical. It's like complaining that someone in a painting looks like you, and you know the painter has seen you before, so the painting must therefore be based on a sighting of you in public, and you demand restitution for unwittingly being their inspiration without giving them your consent to inspire them. Arguing that the painter should have contacted you, and sought permission to be inspired by your public appearance.
Yeah, I already got the impression that you didn't give a shit about the consent side from your earlier posts.
Let's say your comment isn't facetious, and you actually believe this is a reasonable thing to request... You believe that the original LLM creators should have contacted each and every person on the Internet, tracked every single poster down, and got them to sign a consent form allowing the information they had already made publicly available to be used to educate a large language model? You believe that is both reasonable and feasible?
It's the same completely reasonable standards we hold for everybody else? I am just as curious as you are. Does the fact that "'not guzzling all the information and media from people who specifically don't want them to' is such a difficult and infeasible proposition that they can't possibly honor it" not set off any alarm bells for you? morally?
No, because the information does not disappear. It is not gone. It is unchanged. The availability is the same before, as after. The number of people with access to the information "guzzled" is unaffected. I don't know how many different ways to say "it doesn't matter" there are.
I got that much. I'm not sure why you were asking whether I was being facetious or not considering that I initially (and correctly) clocked you as not giving a shit about such consent. And going back to the original post, you really don't seem any different from your typical ai bro in your careless attitude towards the art/textbooks/family photos your ai models guzzle up during traning. Why pretend otherwise?
I was asking if you were being facetious because it's such a ridiculous ask. It would be literally impossible. Large swathes of the Internet are completely random, so there would be no way to know who to even contact. Even if it wasn't impossible from that aspect, many of those you'd need to contact won't still be alive, or in a mental state in which they can even give consent. Ignoring both of those challenges, the logistics of contacting everyone and tracking who had and had not responded would be a Herculean effort. It's just a bonkers thing to ask that I had to assume you were just saying shit to rile me up, because the alternative was not flattering to your mental capabilities.
What the hell is a "typical AI bro" and why do you feel the need to use immature nicknames for people? Should I say "you are such a typical AI hoe" to mean the opposite of whatever you mean?
If the art/textbooks/family photos are unchanged, still available, and in no way harmed, why does it even matter? There's literally no way to know if any given image was even used. If they didn't tell you, it would be impossible to know. Asking for consent for something that doesn't change anything and is impossible to discern even happened is laughable. Why would literally anyone care?
It's like saying you need to get consent from someone to think about them.
I would certainly and accurately use this label for the kind people parroting the same "too difficult" excuse as AI lobby groups for one.
Have they (or you) considered that if it's too difficult to do in a way that respects the consent of the people you're taking from, maybe you shouldn't do it? (Rhetorical question of course, I know you already gave such excuses as "the cat's out of the bag", etc).
Clearly the many creators of the work they're taking from care. Do you only respect consent when you happen to agree with their world view?
Stop reading my messages I've put on the Internet. You're not allowed to get any information from them without my consent. It doesn't matter that it's public information as soon as I post it. You don't have my consent. Why are you still reading? You don't have my consent to read this.
See how stupid that is?
If that's what you want man 😅. I think it's a stupid thing to ask but it should be up to you.
That's not the scenario we're complaining about and you know it.
I was talking about using AI as a tool during software development, always have been. You, and others, are determined to turn it into a discussion about the ethics of data centers. You're not discussing the usefulness of AI as a tool to developers, you're instead distorting the conversation to a tangent, blaming me for it, and you know it.
Same as us. I guess. I've only been following our thread between the two of us, so I can't speak for others. But yeah, that's what I've also been talking about?
No, I'm not, because that's not what we disagree on, first of all... 🤷♂️
Yes, second of all, because those are the most pressing and disturbing and important and urgent aspects in all of AI, the ethics. Of course they are, why wouldn't they be. Nothing is as interesting or important as they are. Nothing!
And I'm not blaming you for anything. No need to play a victim. We're having a discussion. 👍 You're okay.
I don’t really care what Linus Torvalds has to say, unless it’s about the Linux kernel, and I don’t think his opinions matter outside of that. Being a great programmer doesn’t mean he’s also an intelligent, socially aware person.
He did an interview with Linus Tech Tips a few months ago, and when asked about artists being “upset about the large scale theft of work” coming from AI, he said “that’s reality, deal with it” and even went on to talk about photographers that are out of work because “you can fake pictures so much better now” while, within the same breath, saying programmers wouldn’t lose their jobs because he wants to have his cake and eat it too. (Obviously I’m paraphrasing a bit)
I do agree AI can be a useful tool, and it’s a position I’ve held since OpenAI was first starting to become recognizable. But anything that’s generated by AI is slop, and I’ll continue to call it such even if nobody else will.
The movie “Spider-Man: Into the Spider-Verse” used AI tools to place random blemishes and marks (that the artists created) on characters faces across multiple frames. I don’t really feel like finding a source for this one, but I’m sure you could fairly easily if you’re that curious. I think that’s an actual use for AI. Streamlining and speeding up the creation process to give creative people more time to do actual creative work. I don’t think asking an LLM to generate or “create” something for you is anything but slop.
Slop isn’t about quality, it’s about the morality. The, again, large scale theft of work required to get to the point of being able to generate something of a passible quality level, plus the destruction of the environment in the process. I think it’s pointless to reject AI tools entirely due to how prominent and widespread they are, but “creating” something with AI will always be slop.
Source for the LTT interview: https://youtu.be/mfv0V1SxbNA
About 33:10
I'm conflicted about it too. But there is a useful software development tool in there. Its a beige mind good at regurgitating cookie-cutter work. Its pretty good at scaffold and cookie cutter unit tests.
I hate how it makes my colleages behave, but I can totally see the legitimate $500M/yr industry worth of software in there, under all the BS.
There's a Jetbrains or Atlassian size companies worth of product here.
Many people are incapable of nuance.
I think it has some use cases, but it’s like religion: Don’t cram it down my throat, keep it to yourself, and I don’t mind it if it’s on my terms.
Totally get what you mean and I can fully respect that stance. In recent times, it was more like a “let’s bash all the people who think slightly positive about AI”, so kind of the same indoctrination, just the other way around.