China Readies $27 Billion Chip Fund To Counter Growing US Curbs

China is in the process of raising more than $27 billion for its largest chip fund to date, accelerating the development of cutting-edge technologies to counter a US campaign to thwart its rise. From a report: The National Integrated Circuit Industry Investment Fund is amassing a pool of capital from local governments and state enterprises for its third vehicle that should exceed the 200 billion yuan of its second fund, according to people familiar with the matter. Known as the Big Fund, the state-backed firm is expanding its remit just as the US prepares to sharply escalate technology curbs designed to curtail Chinese chip and artificial intelligence progress.

The establishment of a much larger third fund -- directly overseen by China's powerful tech ministry -- signals a resurgent effort to harness the world's largest semiconductor market after years of mixed success with central stewardship. Huawei and its partner Semiconductor Manufacturing International Corp. still had to rely on US-origin technology to build an advanced processor last year.

What a missed opportunity

[Rosco P. Coltrane]

Known as the Big Fund

This SO needed to be called Kung Fund...

LLMs will accelerate China Tech

[laughingskeptic]

The ability to RAG over the terabytes of stolen U.S. corp tech documents is going to greatly aid China in catching up. They will be able to ask an email corpus questions like "Where there any integration issues with implanter technology from X and what were the solutions to the problems?" and be able skip over problems that engineers had to solve the first time.

US campaign to thwart its rise

[oumuamua]

Uhm, that part was not supposed to be said out loud. It makes people wonder why the US needs to do that instead of getting its own act together.

Re: US campaign to thwart its rise

[SuperDre]

The US needed to do that because the were falling behind in using its 'own' created technology. I say 'own', because if you look closely, most of the technology is not even created/invented by an american, but by asian/european scientists. In the end these embargos will cripple the US itself as China will move forward much faster and will not be reliant on US imports, so US companies will miss out on a lot of export income thanx to their own government.

Re:

[timeOday]

Huh? There's never been any secret about the purpose:

https://www.nytimes.com/2022/1... [nytimes.com]

Kinda rooting for China on this one

[supremebob]

I'd imagine that Chinese competition in high end CPU's and GPU's will help to drive down prices.

I'm looking forward to some Chinese upstart breaking the AMD/Nvidia duopoly and making GPU's affordable again. We shouldn't have to pay $500 for a graphics card that can play 3D games properly on a $200 1440p monitor.

Who knows, it might also help to drive down the prices of those new portable PC gaming consoles to something affordable enough to give my kids without worrying about them breaking it.

Re:

[OrangeTide]

China is a centrally planned economy. They are going to kick our ass at these initiatives every time. And they sentence officials and CEOs to hard labors when they catch them in corruption.

But central planning has a lot of drawbacks and they can make bad bets and struggle to adapt to changes in the market. The Soviet Union's administration of a centrally planned economy would be laughable if it wasn't so damn tragic. But then again, the Soviet Union was ran by some of the most incompetent people in history,

Re:

[EreIamJH]

China doesn't have a centrally planned economy - at least not since the 80s. The vast majority of the economy is market based, with little in the way of price control. Most importantly, businesses don't have government imposed production quotas.

What does happen is that Chinese businesses pay a lot of attention to government five year plans. Why? Because a five year plan is essentially a menu explaining where the government will be spending a lot of money over the next decade. Some businesses might see

Good Luck

[Whateverthisis]

That's good news for China!

It most certainly is a positive response given that China canceled the one meeting where they describe their economic results and projections [businessinsider.com] just a few days ago. But I'm sure all of the economic data out of China is solid that this is certainly achievable and not propaganda.

What will that 27 billion buy?

[ctilsie242]

What is this chunk of change going to buy? A bunch of cash to SMIC to build a fab and adapt to 2nm, 1nm and smaller process nodes?

China can diddle all they want with Verilog code, floorplan, and let their Palladium simulator of whatever they are doing run Linux with their new code. However, all that is pointless until not just first silicon is reached, but with a relevant yield count. If the yield count is so poor that only a few chips out of each wafer can be used, and the chips can't be binned, then it won't be viable.

China has a tremendous advantage here, that there is no difference between a company and the government. However, they need to get smart, and use the money for something that will give them a competitive edge. If they spend all their cash on making an AI CPU, but are still on a 10nm process node, it almost is pointless. If they go whole hog and are able to get 0.5nm [1] fabs going, but the process is so expensive that at most, they get a single chip from a wafer, with most of the other items can't be used, even if binned or the bad parts shunted aside via e-fuse blowing, there isn't a point.

Even with both of those, who is speccing the CPUs? You can have a fast as hell matrix multiplication CPU, but what happens when AI tech hits a wall with LLMs and veers into another direction, for example, if GANs or genetic algorithms come back into style, or perhaps some logic added to help see how a LLM comes to a conclusion via some AI-inspired introspection. Even then, CPUs are still needed, not to mention GPUs. AI is nice, but that isn't all of computing. Don't forget quantum stuff.

It is almost amusing seeing how things go on in different places. Here in the US, some private companies (IMHO) seem to want to just use CHIPS act money for stock buybacks, get frustrated that the government won't let them do that, so don't bother with applying for it, claiming "staff shortages", while they lay off large chunks of their workforce.

I do think the US should have just offered to contract out a fab, which would be owned by the government. Let whomever build it, provided they allow the US government to have access to design documents, and once built, had it be used as a fab for fabless companies, or if having the US government make a profit on something is bad, lease operation of the fab out so some private firm gets the cash from it. This way, there is a definite goal in mind, and there can be milestones in place, and the fab is something that can be repurposed for stuff come wartime without anyone hemming and hawing about it.

[1]: At this scale, nanometer measurements can be subjective, so using what everyone uses, like Apple's 3nm CPUs from TSMC.

$27 billion isn't a lot

[backslashdot]

You can pour billions into something but if you don't have the right people -- ie, people who actually give a shit .. it's flushed down the toilet.

Look at how many billions have been blown on full self driving and robotics. There's still no robot that even comes close to passing the robotic Turing test. A test that is not even that difficult (tasks include picking out a rubber band from a bag of trinkets, picking up a small screw from a table and screwing it in -- things that an 8 year old can do in less t