China’s domestic semiconductor industry landscape has changed considerably. The Biden administration has continued to impose export control restrictions on Chinese firms, and the October 7, 2022, package of controls targeted not only advanced semiconductors (such as GPUs used for running artificial ...
It's going to turn out that a nation of billions can, surprisingly, figure out how to make chips domestically once it is no longer possible obtain them efficiently from external markets. This might take a few years to ramp up but it will happen and the market will be flooded.
Most people I've talked to on forums like this believe it is impossible for various reasons that center around technical competence.
I am no lover of Chinese governmental policies but this attitude is both racist and risky. I am old enough to remember people saying Japam was only good at copying Western inventions, for example. I also remember a few years later when everyone wanted to learn to speak Japanese and The Book of Five Rings was the hot CEO book of the month.
Nope. China won't figured it out and will only be able to do semiconductors that trail behind a lot for a very long time.
The smallest node that actually works for them is 16 nm, SMIC's 7 nm in the new Kirin 9000S Chip is not even close, especially compared to the Krin 9000 wich is made with TSMCs 5nm node. Performance and Efficiency are up to 45% behind TSMCs stuff (Which is not something that can be explained away with just because of the smaller node).
But that's not all: The way SMIC manufactures these chips is extremely expensive, due to the fact that they have to use multi-patterning. This also effects yields significant. They have no chance to compete on the open market against Samsung, Intel and TSMC, even with the high subsidies from the Chinese government. Also, while the way they are being produced allows for 6 nm, the gate length and contact width are going to reduce yield even more.
I feel like this is very short sighted. Yes, they can’t do it now. Yes, they are far behind…
But as a manager and a father, the textbook way you get someone to truly learn something and grow is to give them pointers, give them a reason to want to do it, and then let them figure it out on their own. This is how kids learn to walk, how people get good at games, how employees are pushed to learn and grow in their roles, and how countries develop their own tech.
China clearly has enough examples and pointers (legally or not), and now we have a given them a reason to do it (barring them from importing it, but still needing the tech). It will take a while, and their end goals and processes might be different than what ours were. I.e., Sometimes my kid thinks of doing something a different way and it still works. Time will tell. But in the end, they will have their own logistics, their own factories, and their own products. They might be worse, but they could definitely be better, that’s all up to them.
If you wanted China to stay dependent on us, then this was not the right move.
7nm isn't close to 5nm? There's a difference of 45%? What! No way!
Fact is, the Huawei Mate 60 has moved over 30 million units. While that's in no way comparable to Apple's 200+ million units annually, it's a significant scale representing a robust supply chain that's capable of churning out functioning chips. If, by your claims, yields are low because of an immature process, then you'd only expect yields to go up as the process matures.
Yield is not a static factor, but one built on by process development and co-design. You can look at how Intel's yields have increased over the years: they refrained from using EUV on their 10nm (Intel 7) process and, while they ran into a bunch of engineering challenges and delays, still ended up shipping Intel 7 at scale.
These aren't unsolvable issues, but ones of engineering and manpower and skill. EUV still requires multi-patterning for 3nm, so it's not like the problem has been eliminated.
That is the opposite conclusion of this article you didn't read and it's extremely unlikely for many reasons.
China will eventually be able to compete in the mass semiconductor commercial market fort non-essential chips, but are about as far away from today's advanced chip manufacturing as your '99 iMac is from the Oculus Rift.
If the sanctions remain in place and are effective, the only way China will catch up is to develop an entirely new manufacturing process(which they are trying to do, blindly) as well as develop comparable physical fab technology to tsmc(which at their current tech is like trying to build a pixel 9 from scratch with hands tools).
Anecdotes from forum users can be entertaining but are not indicative of anyone who knows anything about the industry, and this article explains very clearly why the market will change but China is facing near to completely insurmountable challenges when it comes to competing in the advanced semiconductor market.
Taiwan, the only country that could help china with this, is strongly opposed to China on most fronts, direly opposed to helping China technologically since that would mitigate Taiwan's leverage, and has allied themselves with the most powerful and technologically advanced country in the world(outside of semiconductors; smooth move Taiwan), another opponent of China based on their national security, on which the states spends obscenely more money and resources on then anyone else.
"Well china could take over taiwan our they did Hong Kong."
In which case, tsmc would physically destroy their fabs and data, China would gain nothing and lose any goodwill they've been cultivating and spending billions on, with more sanctions than before and end up even further behind than they already are.
The distance China is late to the chip game and the extreme technological limits of cutting-edge semiconductor manufacture are not racist and it's nonsensical of you to imply they are.
It's not impossible that the Chinese could catch up eventually, and they have plenty of resources to try, but manufacturing advanced semiconductors without knowledge, resources and access is not an obstacle to be overcome like the Japanese refining manufacturing methods for mainstream global technologies and factory management strategies, this is more like trying to throw a dart without any hands or feet from your apartment to a dart board in another country, and then the dart changes into a butterfly, lands on a flower, pollinates it and the flower turns into a baboon that composes sonnets.
Frankly, I think you're missing the forest for the trees.
Yes, SMIC is behind the big players, but these newer node shrinks are delivering increasingly small iterative gains. Meanwhile, leakage power is becoming an increasingly important concern for smaller nodes, and much of the technologies that these advanced nodes are delivering (e.g., backside power, 3D stacking) are not coupled with the node itself. Indeed, there's no reason GAAFET (or whatever people are calling it these days) needs a smaller node, only that it doesn't really make financial sense on a larger one. It would, if, say, you wanted to develop the technical capability and had idle engineering cycles waiting for machines. So, what does this mean? SMIC can progress at will on everything except the node itself. It'll remain behind on sheer density, but Intel showed that you can be stuck on a single node for the better part of a decade and still be a market leader. SMIC already knows how to scale a fab (the Huawei Mate 60 moved more than 30 million units in less than a year).
On the process side, SMIC will be behind but not obscenely so. What about at the architecture level? Here, there lies a big problem. Most of the world has been locked in to the CUDA ecosystem. They're tied to Nvidia for any massively parallel computing needs, and Nvidia can charge whatever price they want. This is, unsurprisingly, a problem, because Nvidia's hardware is, believe it or not, not the most optimized for a wide variety of applications. There's been an abundance of research into specialized accelerators for applications like machine learning and scientific computing (and indeed, also many real-world designs) which deliver up to an order of magnitude increase in perf/W, but, perhaps more importantly, a substantially reduced TCO due to not having to pay Nvidia's obscene prices. Indeed, this can be seen by the fact that Huawei's Ascend accelerators are actually getting tractíon in the Chinese market. So, at scale, China will be hogging up more electricity for low-capital-cost data centers with higher operating cost... Fortunately, China has an obscene amount of basically free green energy coming online in the next few years.
So, what's all this concern about? It's pretty simple: mobile applications are constrained by energy-efficiency, and you can't get around key energy-efficiency limits without shrinking the transistor. Fortunately, advanced nodes aren't used in most military or space applications due to reliability concerns... So, the main application for them will be consumer-focused applications like smartphones, autonomous vehicles, and drones. This is a rather annoying problem given that China is betting their next decade of growth on the EV transition, but it's not insurmountable in the near-term given the sheer cost advantage China has in shipping non-autonomous EVs due to efficiencies in the rest of the supply chain. The same holds for drones, where DJI is the undisputed market leader. For smartphones, Huawei has the largest captive market in the world and Apple is rapidly losing market share.
Ok, so, having established the challenges, how far away is SMIC from a solution?
As discussed, SMIC already operates fabs at scale for 7nm, and indeed SMEE has demonstrated the ability to ship a DUV lithography machine. Intel7 is also entirely DUV, so this isn't all that surprising. NAND processes still use DUV as well. Intel's TSMC N7-equivalent process first started shipping with Ice Lake in late 2019, while SMIC's 7nm process started shipping with the Kirin 9000S in mid 2023. SMIC is expected to start shipping 5nm at volume later this year (how? I'll never know, but clearly the yields are alright because they're going to be supplying Huawei again).
I'll stress this again and again: yields are for the most part, an engineering problem. Getting high yield with multi-patterning requires some absurdly complex engineering, but it's not inherently impossible with tight enough tolerances. It wouldn't be economically viable in an open market where EUV is available, but it isn't.
The big concern right now is, surprisingly, not on the manufacturing side but on the software side. There is no equivalent in the world for American EDA tools. None. The minute sanctions stretch towards EDA, China will be set back by more than a decade. China lacks EDA tools for newer nodes.
China will eventually be able to compete in the mass semiconductor commercial market for non-essential chips, but are about as far away from today's advanced chip manufacturing as your '99 iMac is from the Oculus Rift.
Their 3.7Ghz x86 CPU on DDR5 is only about 7 years out of date tops only on the clockspeed. Kinda close to a coffee lake.
Not to mention China has been funding and driving RISC-V research and development for years, especially with their open source projects using linux.
They're not doing this blind, even 26nm die manufacturing is common knowledge now.
High end chips just happen to be a very expensive market both in resources and technological scale, which is why it primarily grew as a global consortium of companies relying on each other to produce the final product
China has already perfected the art of replicating a globalized industry at home because they have the resources and internal funding. The complexities and proprietary information involved with current gen chip production won't stop them lol. They don't need to aggressively attack Taiwan in any way to achieve what they need. Maybe some cyber espionage to speed things up, which they have done before with military tech, but nothing more lol.
This is a very different take from "Chinese people just can't do this no matter how much time, money, and talent they devote to it" which is what I'm taking about.
It's absolutely incredible to me that US thought they could hobble a nation of 1.4 billion that has a superior education system and produces more STEM graduates than US does. This was the definition of hubris.