Today I opened newegg for the hell of it and figured I'd look at the cost of a 2 TB NVMe ... Fuck man. This is seriously depressing.

BUT! I feel that lost hope could/will turn into a new forms of "min-maxing" or whatever where we will see people focusing on small, efficient software to make the most of what we have to weather the drought. We got spoiled with cheap hardware. Time to make the most of all that cheap hardware we have.

Though of course, replacing this hardware as it breaks down will be expensive. Perhaps we could go back to 1/2/4 GB RAM sticks and 256/512 GB Disks. Then run minimal Linux installs, minimal GUIs, minimal services, minimal software.

To amend this: Reading the 8088 domination article and the little $8 RISC-V Linux board I think we have plenty of opportunity. Maybe we learn to shed our inefficient BIG computers for small, low power, purposeful, yet general computers. I play with Plan 9 a lot and its portability, small footprint, low resource, and distributed design has showed me we can do a lot with a little as long as we stay disciplined to keep things simple and to the point.

I think there is a second order effect of this: it won't be worth making games for PC. At least, large budget games.

I'm still hoping some future LLM is good and small enough for people to run local AI for everyday life.

There a lots and lots of great games that will run on computers that would be considered low power today. And I’m not even talking about older games.

It's also killing the console market. Sharp drops in all console sales after "adjusting" to ram prices.

I agree it’s brutal right now.

I have to think, it can’t take much time before the Chinese DUV fabs come online and non-frontier chip prices crater. Sure, you might not be able to get 5090 equivalents, but the 30xx series was DUV and so that should be replicable in a few years.

Whatever quality bar / process node China gets to you know they will scale it out super hard.

(Some quick investigation suggests that the Yuliangsheng DUV machines currently under construction are 28nm, which would map to 2014’s GTX 900. Perhaps I’m being overly bullish but I assume they will recapitulate/clone the tech faster than the rate it was originally discovered.)

Its interesting because I was at Costco the other day and they weren't even selling PCs unless they were gaming ones. It seems most consumer electronics have switched over to laptops.

Also, this reminds me so much of the download more RAM ads back in the day.

Sure they will. The same components go into every office PC. As for GPUs, we could always play older games.

intel makes gpus :). Maybe games will be finally made again in a way that a PC can run them without needing a 3k GPU in there to get toward half the refreshrate of your monitor.

gaming does not equate to only things like the last of us or cyberpunk which need extreme machines. a lot of visual fidelity can be reached still without having such a monster. Just need to not use certain features which are currently pretty much exclusive to Nvidia...

At least you'll be able to get a Pi 5 for $700.

If nobody can afford a gaming PC, devs will need to make games that run on older hardware. This isn't a bad thing.

On the bright side, you’ll be able to play Skyrim-halo-Pokémon-Minecraft as much as you want.

If the AI bubble bursts and prices normalize because demand for AI hardware is reduced, then these companies might pay more attention to other markets like gaming again. But who knows the future anyway.

What is more sensible: Listening to the best music on an average sound system or listening to average music on the best sound system?

There should be a big enough back catalog of great games for you to enjoy without having to buy a new gaming rig.

I'm just sticking with Tetris.

Regarding graphics capabilities, haven't we reached "good enough" like a decade ago? Does anyone actually want to see ultra realistic gore when attacking an enemy? I'd also submit that photorealistic art style is not the most popular one either nowadays except for a handful of special sims like race car driving.

Not only if the internet dies. If the country or countries managing the AI decide they don't like your country and ban you from using their (major) AI providers, in a world where most knowledge and work tanks will require access, you're screwed.

Not even talking dictatorship scenarios here. The USA might simply decide it doesn't like your country. Or maybe heavily cripple the models you can use, or censor them, for whatever reason.

I guess this is an argument for local AI. Probably not feasible for the less technically savvy.

- Full time professionals do a better job of maintaining a reliable and secure service than end users

Also:

- local latency is lower than network latency

- latency is important for action games

One consequence has been lower power requirement for local machines. This is partly why a 2015 MacBook is still perfectly serviceable for basic use in this day and age.

The AI boom first reversed the trend, to now try to get back to that cycle.

People long ago stopped buying computers and started using Saas on their mobile devices instead. Ask HP.

that was when RAM and GPUs were cheap. Now that is dying, too.

Can you think of examples of this cycle? I know this is roughly what happened with mainframes, but I struggle to think of other times it went like that.

Maybe arcade games?

There's also major not-startups working on it, such as Broadcomm: https://www.broadcom.com/info/ai/3point5d

Supercomputer do fit in your desktop though. Most supercomputing challenges of the 80s or 90s fit comfortably on a modern home pc.

Propaganda would still propagate through more closed networks. There's propaganda on TV. And at the bus stop.

> And your local AI could emulate the good parts of the internet.

Like talking to people?

Trump and AI-power security research are both doing a good job of promoting installable software!

Which is not a bad thing per se: most of use could use our devices a little longer than we usually do.

most games which require that big PC are poorly optimised and often times they arent even fun. these companies focus on visual fidelity and give all their engineers insane machines to build the games on. that has beenn destroying PC gaming way before AI/LLMs came into the picture.

I am using AI now to make a game engine that will run and make 'nice enough' graphics on simpler machines. things like laptops with iGPU and 4gb of ram.

U can still get quite ok graphics, and make very complex and rich gameplay systems using such a machine. It just needs to be the target machine while developing so the right tradeoffs are made.

Why would component vendors allow shortages to continue indefinitely for the benefit of hyperscalers and AI companies, especially given what may be years of pent-up demand in other markets?

And why would the game industry, considered as a whole, want to incur higher costs to exclusively deliver its product in a way that provides less value to the majority of its customers, many who won’t be willing to pay the higher prices required to maintain similar profits?

Recompilation at most threatens platform exclusivity as a business model. Otherwise, it’s just piracy with extra steps.

At most, I expect to see already existing trends towards more live service components and microtransactions continue, with digital delivery only and a required online component in many cases, but with most if not all of the actual game code and assets hosted and running locally.

As for non-gaming applications, content creation and a software development are two examples where "cloud native" workflows remain cost-ineffective in many market segments in spite of hardware price increases, and will remain so as long as cloud providers continue to nickel-and-dime customers for IOPS and compute.

I don't know how much of a conspiracy that is. If the current trend of hardware costs keep rising to the point of pricing people out of having any real compute locally, and chip makers actively de-prioritizing consumer-grade memory, GPUs, etc, companies will (more) heavily push thin clients and streaming for everything.

I would have laughed at the thought a few years ago, but now I am slowly starting to adopt some of the viewpoints of American second amendment hardliners.

It's just unfortunate that producing a half decent computing platform is significantly more difficult than a rifle (and that in the case of said hardliners, the capital is backing them, here it's working against us).

> A sensible point of entry is a 2899$ Mac mini.

> If you're not partaking that's only on you.

There's honestly nothing sensible about that for a lot of developer budgets across the globe. Don't assume most folks have $3k to burn.

Yes, but that's a lot of money if you don't already have the hardware, plus you'll still be paying for electricity costs. (No comment on relative capabilities.)

$3k being a point of entry for a pale imitation of the models being compared to is not very inspiring.

Surely this is a joke or sarcastic comment?

They absolutely can't do what, or more of what the biggest models can do. There's nothing frontier about small local Qwen or Gemma models.

2899$ gets you a machine that is just good enough to run some small models, slowly, with terrible context, requires a ton of tinkering and fine tuning, managing and constantly downloading a bunch of huge file blobs, etc.

Partaking those $2899 in most of the large cloud models would bring you much more productivity, much less boilerplate and time spent in setup, etc.

So.. just weird accusing people here of "not partaking" in AI when you're just essentially toying around if you're actually comparing local Qwen and Gemma's with frontier models.

Yes but then you'd have to give closed-ecosystem-ass Apple money, which is worse than giving it to Nvidia who at least support linux properly. And you still get prompt processing speeds of a leisurely sloth.