Eastern China or westernmost state of the USA? You decide!
AlteredEgo
joined 2 years ago
Afaik Chinese DDR4 ram production is already keeping prices in that segment sane. They are working to break the ASML monopoly and sanctions on extreme violet light lithography to make the most advances chips and then will ramp up production. They are churning out like 10 times more engineers than the US and EU. They are ramping up but not sure when they will be able to compete at the highest level. I read 2030 earliest.
Production of basic necessities and planning an economy is good. China build so many apartments they now have 90% home ownership instead of treating housing as a speculative asset. Now they caught up in high tech and in will be producing highest tech computer chips. And they are building nuclear power plants so you can actually buy AI tokens that do not contribute to the destruction of our world.
Also North Korea is far less dire than depicted these days. I wouldn't want to live there but life expectancy is something like 74 years so overall it's probably far from the worst country on earth.
You shouldn't trust the mainsteam western media or institutions about any of this.
The US is ruled by a junta of oligarchs like the Nerd Reich techofeudalists backing Trump. They believe in nothing, only wealth and power. Trump is no King, but a puppet and the US empire has competing interests.
Xi Jinping is similarly not ruling like a king or dictator. He is accountable to 1.4 billion people and part of what drives his party is ideology. THEY BELIEVE IN SOMETHING. That alone makes them far less evil. They are not benign but they do have a track record of improving the material conditions of billions of people, less predatory economic imperialism and no foreign wars or interventions or coups destroying other nations.
There is no doubt which one is more evil. China now leads in like 58 of 65 advanced technologies. Lets just hope the remaining socialist idealism holds and they stay less evil.
Obviously fuck Meta, but the difference is the publishing. Scraping and copying for personal or business use is a civil matter. And machine learning from unlicensed material is also fine - as long as the material isn't "memorized" and an AI model can't reproduce it.
Aaron bravely published the papers which is a criminal matter. That is why I believe copyright and IP law is the real theft, they take down copies for people to learn from.
There are many many millions of people who read and learned from pirated textbooks and who use those skills to do things. Who watched pirated amines or comics use that to learn how to draw.
Basically we should not be siding with the unethical side of IP law just to oppose AI corporations. They can afford to buy the books and media, and a simple purchase will do. And they can afford the lawsuits. And they will figure out the memorization problem, so that AI models learn but not memorize (which is only happens in like 1% of the cases and only when you specifically ask for something "just like that").
China (so far) is saying that the AI models they produce should be open weight and be available to all people on earth. So if we have any issues it should be with the monopolization of AI models that concentrate this new developing immense power of AI in the hands of a few plutocrats. Which the IP laws might actually help them with.
The billions in funding for the AI corporations also went into B2B marketing, presentations for businesses and schools etc. A problem of neoliberalism / capitalism. The techbro half of AI psychosis is deliberately inflicted.
And the US / western economy collapsing isn't only related to AI though. When the bubble pops there would have been massive layoffs anyways. The techno barbarians behind Trump literally want to dismantle and destroy the US to create their own fiefdoms from the rubble: The Nerd Reich”: Author Gil Durán on Big Tech Fascism, Peter Thiel, JD Vance & the War on Democracy - Lemmy. So DOGE, the tarrifs, even Iran shutting down the oil supply is all desired by these people.
If we don't solve the cause of these symptoms we're not going to fix it.
I couldn't find anything about concurrency. But I assumed it's more like ~1000 users? B300 has something like 100,000+ tok/s, not sure about context size. I also was wrong on my math for the solar panels, see above. It's more like +50% costs than 10%. Obviously all of this is napkin math, but even 50% isn't too bad. My point is that with a slightly saner rollout of AI there seems to be absolutely no problem to make it sustainable. And if you buy your tokens in China you can select a data center that gets it's power from nuclear or solar.
And yeah we don't know how long they will release open weight models. Once hardware to run such open weight models become available they might change tactics. But right now China releasing these open models has a serious impact on the stock price of these AI companies, which is a good thing. The sooner the AI bubble pops the better.
And like 7 of the top 10 suppliers of tokens right now is from China apparently. What Xi himself said is that the world needs a multilateral approach and AI should not become a monopoly. And while that is clearly propaganda aimed at the global south, China does have a different approach. They spend peanuts on their AI models compared to the west (and the algorithmic breakthroughs done by Chinese AI researches afaik have also been published).
But you actually can use an existing model to generate training data for a new model. Which is what the US corporations are currently accusing China of doing, which is of course hilarious. Of course China has generated huge amounts of training data themselves. So basically any industrial country could in the future fund a research effort to produce a new AI model and make it open weights. A monopoly situation would be another huge issue with AI.
Basically in regulation, sustainability and democratizing access China is doing very good. And in poking the AI bubble.
A single DGX B300 that could run Kimi for a few concurrent users, uses 14 kW.
Huh? 1400 TDP = 1.4 kW. This is official I believe.
And maybe my sources are off but: "As of July 2026, a single B300 GPU purchased outright runs about $53,000 (Spheron Network, July 5, 2026)" https://tech-insider.org/nvidia-blackwell-gpu-pricing/
Uh you're right, you need about 4x so 270 kWh over the day, which requires about 270 / 6 * 24 = ~1000kW. With about 175€/kWp panels that is another 175 thousand plus installation. And something like 4000m². But that is all the energy costs for 30 years up front. The bigger constraint is the land use then.
To make AI data centers sustainable with solar and batteries costs about 10% of the price of the base hardware. The AI hype is still insane but we can have a little AI.
Obviously the AI rollout should be slowed down and AI use regulated, like stopping any replacement of workers by AI.
But China is set to win the AI race with their open weight models on par with the current frontier models. And they are building nuclear power and solar and wind like no tomorrow. And they just started to produce sodium ion batteries to solve the grid storage problem.
Spurned by the embargo they are developing more efficient algorithms and "Extreme UV lithography" to catch up to the embargoed chips. The neoliberal model of the west is now falling behind even the high tech sectors.
In case anyone is curious it takes 8 Nvidia B300 cards to run this Kima K3 frontier model yourself (2.8 trillion parameters), costs about half a million. For a big corporation or institutions even in the global south that is relatively cheap and that is the top tier option for running a frontier model. Including complete data sovereignty. Using AMD will be cheaper, and prices will go down once China catches up in like 2030. And you could run that on solar panels and batteries and regenerative water cooling. 8x 1.4kW x24 so maybe 300 kWh LiFePo4 battery so another 50k. ~~To power one complete AI server you'd need about 134 m² in solar panels which are basically free these days.~~
For solar panels you'd need about 270 / 6 * 24 = ~1000kW. With about 175€/kWp panels that is another 175 thousand plus installation. And something like 4000m². Plus chargers, and large public pool to cool those 270kW. But that is all the energy costs for 30 years up front. The bigger constraint is the land use then.
So basically less than ~~10%~~ 50% extra costs upfront to make AI data centers sustainable. This is a capitalism problem, both the hype / bubble / psychosis (b2b marketing induced) as well as the sustainability question.
Even Linus is conflating “doing more” with “being more productive”.
No he isn't:
Torvalds’ broader argument is that generating code has become easy, while maintaining a system over years or decades remains hard — and AI does not close that gap on its own. Without a working understanding of the underlying architecture, he suggested, AI assistance can help someone make mistakes faster rather than build something durable.
They are working on cracking EUV lithography. From the sweatshop of the west to leading in 58 of 65 advanced technologies already. Neoliberalism has been the ruin of the west.