today at 9:56 PM
Even though it only runs qwen 3.8 27b at a 2 bit quant, it's still highly viable, especially for things where you just need a constant token stream of slightly below average intelligence, ie this would be great for things like:
- monitoring server error rates
- reviewing security footage
- watching plant growth / controlling watering
etc. Any long running task you can now just slap a PS5 on it. 70 cents a day in power cost means you now get ~2M output tokens per day (~$3.50 worth of usage). It's not the worst thing in the world if you have one lying around. 2bit isn't even that terrible - it's ~80% the accuracy of 27b at full quant, which should put it at around GPT-5 levels of intelligence (which honestly is crazy in its own right).
today at 8:26 PM
I've never even considered getting a PS5, but hackers converting a locked-down appliance into a general purpose computer has me really considering.
today at 8:39 PM
Well that locked-down appliance is way under-specced compared to an equivalent build at the same price.
today at 9:11 PM
Does that remain true in the RAMpocalypse?
today at 10:04 PM
A PS5 is $600. This random mini PC with the same amount of RAM is $385. https://www.bee-link.com/products/beelink-eq14-n150
I'm sure someone will point out some spec that's better on the PS5, but the extra $200 will make up the difference if you spend more time looking than the 15 seconds I just did.
(or if you just build a PC from parts yourself)
today at 9:18 PM
Even at the current increased prices - you can build a PC with a better spec and with a 4K Blu ray drive for £569? I'd be curious how.
today at 8:05 PM
The project's primary site should not be GitHub, and I am serious. Playstation 5 is not some historical curiosity and Sony might strike the page down any hour now.
Those projects should default to hosts less geared toward corporate governancy; custom internet domain being the first requirement.
today at 8:22 PM
I'm sure Sony has an opinion of the jailbreaks that enable this, but they really don't give a shit about anyone running llama.cpp. Why would they? They're just glad you're not pirating games.
today at 8:30 PM
Is there a real or imagined hardware subsidy on the PS5 hardware though? In other words, is the current MSRP below what Sony would charge in a hypothetical world where they weren’t involved in any form of gaming software business, with an assumption that they will make additional margin dollars from game sales? I can see some manager making the case that repurposing the hardware for non-gaming use is taking food off his family’s table, in a manner of speaking. If so, it’s an argument that you don’t want this type of use to become popular and have PS5s start flying off shelves and never having games bought for them.
today at 8:23 PM
There's no feasible way to avoid takedowns/DMCA in 2026 unless you do TOR shenanigans. Even with a custom domain, Sony could just DMCA the web host for the project.
today at 9:07 PM
Magnet link, once the DHT networks picks it up (and there are seeders) no one will take it down unless they go after every single IP address
today at 9:07 PM
Torrents. Got the leaked GTA6 videos by torrenting and they're still easily available on public trackers. Don't see why projects like these couldn't be distributed similarly if shit hits the fan.
today at 8:28 PM
Sure you can. This repo has definitely been vacuumed into the latest training data dump, so the next generation of models will happily spit out a similar reproduction, Sony be damned!
today at 8:35 PM
Gitflic.ru begs to differ.
today at 9:04 PM
today at 8:40 PM
> Those projects should default to hosts less geared toward corporate governancy; custom internet domain being the first requirement.
I struggle to think of anywhere on the entire open internet you could put something like this where it could not be taken down.
today at 8:14 PM
Conceivably this should also work on the BC250 cards, which are ex-cryptocurrency mining motherboard built around AMD's "Cyan Skillfish" APU which is a cut-down version of the PlayStation 5's custom silicon. You can still pick them up on ebay for $150-$200.
today at 8:25 PM
Good luck finding one for $150 or even for $200
And at that price point or a little bit more you might as well just get an RTX 3060 and save yourself a lot of hassle (you also get DLSS, DLAA etc.)
today at 8:29 PM
I paid $150 shipped a few months ago then maybe $40-50 for all the parts to run it. The RTX 3060 sells for $400+ in my market and still needs a computer to attach it to!
today at 8:39 PM
Just opened up eBay and they're going for like $200 here, $220 if you want a "modded BIOS" (whatever that means).
today at 9:08 PM
Flashing the bios takes ten minutes with a thumb drive. It’s hard to get wrong but guess some people would be nervous enough about that to pay $20 extra.
today at 10:12 PM
I also figure the seller was just lazy to change it back.
today at 7:57 PM
With this in mind, having LLM-powered NPCs would be a beautiful thing. There's been limited testing on PC games (PUBG Ally), but I can imagine it will be the standard.
today at 9:12 PM
After all of the agonising about AI companions, we realised that we just wanted someone to play games with.
today at 8:15 PM
Fortnite Darth Vader already did
today at 9:29 PM
Kinda cool but also entirely useless given the limited VRAM.
The readme oversells it by pretending it would be something actually usable, when instead it is a cool achievement to make it happen.
today at 10:05 PM
A 20 tps Qwen3.8-27B at 2 bits quantization is limited but not useless. I've been putting one on "read this codebase and figure out XYZ" for me, where the typos are less consequential.
today at 7:04 PM
Not long until Anthropic's dreams of self-replicating models starts taking over PS5s!
today at 7:15 PM
Haha, I was thinking the same, I think we will see the most random devices being able to host models in the near future. An army for the techno core.
today at 7:15 PM
Relevant: https://en.wikipedia.org/wiki/PlayStation_3_cluster
today at 7:21 PM
Obligatory... "Can you imagine a Beowulf cluster of these?"
today at 7:36 PM
Those were relatively big for scientific computing in the PS2 days
https://www.ncsa.illinois.edu/2003/05/27/playing-the-superco...
today at 8:04 PM
For somebody who hasn't kept up with all the different open source models and what token rates mean, how would using this compare to something like codex or claude? I assume it would be much slower and less capable..?
today at 8:12 PM
Significantly slower and less capable than opus 5.5
However you can run comparable or better models than gpt4. Try to stay under 30b
today at 9:32 PM
GPT-4 and Opus 4.8 were the models that started the AI craze. Now people look down on them.
today at 7:07 PM
today at 7:07 PM
Don't give Sony ideas...
today at 7:02 PM
11.4G of shared memory. Could do something interesting if you don't have a GPU.
today at 7:19 PM
today at 7:29 PM
The toaster is next!