Wow. V100. That brings back memories. Way to go.
I put a datacenter GPU in my gaming PC
51–60 of 199 posts
Re: I put a datacenter GPU in my gaming PC
#52> The compute is still real. The VRAM is still real. And the memory bandwidth is where it gets genuinely surprising.
sigh
Re: I put a datacenter GPU in my gaming PC
#53Based on the title I was really hoping to see how this was used for gaming, but they just ran an LLM on it
Same. With no new NVIDIA gaming GPUs this year, seems like an interesting problem to solve.
Re: I put a datacenter GPU in my gaming PC
#54Based on the title I was really hoping to see how this was used for gaming, but they just ran an LLM on it
I don't think that is even possible, every piece of silicon on that chip that is required to do gaming is ripped out in favor of more compute cores.
Re: I put a datacenter GPU in my gaming PC
#55Some resell group is going to have to make this easier. The shear amount of these cards otherwise heading towards the landfill is staggering. That is if Big Tech don't destroy them to prevent model weights from leaking.
Isn't this the same thing with 32 GB already on a PCIe socket?
Re: I put a datacenter GPU in my gaming PC
#56The real question: did your local LLM write this post?
There are many tells aren't there? There was clearly hard human work and experimentation here, but it's a shame the OP let AI do chunks of the writing. Once you see it, it's much harder to take the post seriously.
Re: I put a datacenter GPU in my gaming PC
#57Some context: - In 2017, the v100 was a ~$10,000 GPU. I believe there was a PCI-e version but this is probably so cheap because SXM2 is going to be harder to use; - A 5090 has 1800GB/s of internal memory bandwidth (compared to 900GB/s in the 9 year old GPU). Of course a 5090 is substantially more expensive; - A 5090 has ~21k CUDA cores vs ~5k; - The current $10k NVidia GPU is the RTX 6000 Pro w/ 96GB of VRAM. It has…
> Consider this: in 5-10 years, the trillions spent on AI data centers will likewise be sold for scrap most likely. That's how short the runway is for OpenAI and Anthropic to recover that investment. Even more interesting: it'll devalue all of SaaS and the entire US tech sector. We might have just shot our most valuable non-AI tech products in the foot.
How so? I understand that flooding the market with physical goods will reduce prices and thus profits. But how would that also reduce the nonphysical SAAS stuff?
Re: I put a datacenter GPU in my gaming PC
#58> And yes, if you want the absolute best, Opus 4.8 exists. It also costs more per 20 minutes of heavy use than I paid for this entire GPU and adapter setup combined. But the gap is shockingly small. I don't think this is a fair characterization of the situation. I use frontier models via API pre-paid tokens every single day, and I can barely rack up $100 per month . The fact that we figured out how to burn double thi…
I use hosted providers myself, but I can churn through $100 worth of tokens in half a day even with cheap models like Deepseek easily. If someone's use is as light as yours, then sure - grab a subscription and you'll save far more. For higher use it will come down to how cheap your electricity is whether it is worth offloading at least some of it (for me it's not, FWIW)
Re: I put a datacenter GPU in my gaming PC
#59> And yes, if you want the absolute best, Opus 4.8 exists. It also costs more per 20 minutes of heavy use than I paid for this entire GPU and adapter setup combined. But the gap is shockingly small. I don't think this is a fair characterization of the situation. I use frontier models via API pre-paid tokens every single day, and I can barely rack up $100 per month . The fact that we figured out how to burn double thi…
Claude is something like $35 per million tokens. If I was using API pricing I could trivially spend $100 in a single hour long coding session, with /fast turned on in about 10 minutes. Not sure how you guys are using it.
Re: I put a datacenter GPU in my gaming PC
#60Some context: - In 2017, the v100 was a ~$10,000 GPU. I believe there was a PCI-e version but this is probably so cheap because SXM2 is going to be harder to use; - A 5090 has 1800GB/s of internal memory bandwidth (compared to 900GB/s in the 9 year old GPU). Of course a 5090 is substantially more expensive; - A 5090 has ~21k CUDA cores vs ~5k; - The current $10k NVidia GPU is the RTX 6000 Pro w/ 96GB of VRAM. It has…
> Consider this: in 5-10 years, the trillions spent on AI data centers will likewise be sold for scrap most likely. That's how short the runway is for OpenAI and Anthropic to recover that investment. Even more interesting: it'll devalue all of SaaS and the entire US tech sector. We might have just shot our most valuable non-AI tech products in the foot.
> We might have just shot our most valuable non-AI tech products in the foot.
Counterpoint: the fiber buildout during the dotcom boost. That crashed the economy pretty hard when the bubble burst, but we are still benefitting from all the dark fiber that was arranged for and built out back in that era. A lot of today's ISPs were able to grab up that fiber after the bust for cents on the dollar.
Assume that OpenAI and Anthropic go bust, which at least one of them likely will, and possibly a fair few of the datacenters that are under construction will also collapse. Someone will be able to snatch these physical assets again for cents on the dollar and run open-weight models on them or train new ones.
The problem isn't (and no, this is not an AI tell, everything I write here got typed on a 2022 M2 MBA by hand) the assets, they will be put up for productive usage, just as with any other large bankruptcy or bubble in history. The problem is the "IOU" that is being passed from one hand to the next like a hot potato. Assuming a recovery of, maybe, 20% after the collapse, at 1.6 trillion dollars of assets under management by some kind of private investment/debt we're looking at about 1.3 trillion dollars in valuation that is going to be wiped out.
And given that a lot of the investment market is actually backed by pension funds... this is going to be a bloodbath. Not only will there be a lot of people laid off in addition to the layoffs we already saw "due to AI", but when the pension funds and thus their payouts collapse? We'll see retirees flooding the employment markets who just try to make a living, rendering the situation for everyone else even worse. Flipping burgers used to be a gig for students, these days students compete with people of all ages desperate to survive - and thus desperate to undercut others in wages.
Another problem will be the capacity buildout in the semiconductor industry. It's already heading toward an oligopoly after numerous boom-bust cycles: you only have two and a half GPU chip vendors (NV, AMD, Intel), two vendors of general-purpose CPU vendors (Intel and AMD - I exclude Apple because they do not sell their CPUs to any third party and ARM because 99% of non-Apple ARM chips do not go towards servers, desktops and laptops), three RAM manufacturers (Samsung, SKhynix, Micron) and two and a half physical chip manufacturers (TSMC, Samsung, Intel). When the AI bubble bursts, it will be one of a hell of an effort to prevent at least one actor from going bankrupt.
[1] https://prospect.org/2025/11/19/ai-bubble-bigger-than-you-th...