Chinese semiconductor industry

Except there is no good reason why China can't do both. You are running into the same folly Japan did with their 5th generation computer project. That was when the Japanese semiconductor industry started going off the rails and started on a losing path they would never recover from.


It is a big mistake to ignore classical computing since more often than not when someone comes up with a new computer architecture, someone else comes up with a way to emulate it on classical computers which turns out to be cheaper than the said computer architecture.

gadgetcool5 said:

China should focus on quantum computing, because it is less than 2 years behind the West at most. This area has far greater potential rewards for China than traditional EUV semiconductors, let alone DUV. It is far easier to get ahead in a new and emerging area than an already established area. By 2030 China could easily be ahead of the West in quantum computing.

If you look at 5G, the reason why Huawei was able to get ahead is because telecommunications technology was still in its infancy then. There was uncertainty in standards between CDMA and GSM, the US bet on CDMA and lost. Hence Huawei had an opening.

Similarly, currently the opening is in quantum computing. It's not that China should not invest in traditional semiconductors, but that even with massive investment it may still be playing catch up in 2030. You really want to be ahead, not playing catch up. The best quantum computers in the future will run laps around the best traditional semiconductors.

Click to expand...

Not all computing problems are faster on quantum computers. A very simple example is addition, where quantum computing is actually exponentially slower than a classical digital system.

People seem to have this incorrect notion that quantum computers are just faster and better at everything.

Like @PUFF_DRAGON said quantum computers are specialized machines for solving a certain class of problems. There is no evidence they will ever become general purpose. In a similar fashion the Japanese 5th generation computer effort was hyper focused on logical solving as a problem and was oblivious to the major computer revolutions in the 1990s. All of which were obvious in retrospect. The Internet was nothing new, neither was hypertext. I will leave you with a quote from science fiction writer William Gibson. "The future is already here, it's just not very evenly distributed".

Why did the Internet take off while logical machines lost out in terms of mind share? Productivity gains. I remember hearing about logical machines in the 1990s but it was always relegated to niche applications, like Prolog for robot control, and even there it lost market share. For all the massive effort spent on logical computing it never grew out of niche applications. Prolog is still interesting but other than for solving logical type problems it has little use and you need highly specialized training to know how to use it.

I will add that the existing semiconductor industry has tremendous inertia. You have to remember that you don't have to exceed current classical computer performance with your special purpose computer. You have to exceed it when you get into mass market against the classical computers available then. This has been the cause of failure of all sorts of technologies like Bubble Memory, 3D XPoint/Optane, Thinking Machines, etc. D-Wave comes up with a hugely expensive "quantum" computer which solves simulated annealing type problems. Then someone shows there is a classical computer algorithm which solves the same problems in a classical computer you can buy for peanuts. Like 1/1000th of the price at similar operating speeds. Or much like Ageia and their PhysX physics accelerator cards, the application is really niche, and it turns out you can do it in software easily in some other already available hardware like GPUs. Intel has come up with CPU machine extensions which accelerate it and Ageia didn't use them on their non-hardware accelerated software layer for cross compatibility specifically to make their numbers look better. That kind of thing. The "magic" only lasts until they try to sell it and few fork out for it.

PUFF_DRAGON said:

Not all computing problems are faster on quantum computers. A very simple example is addition, where quantum computing is actually exponentially slower than a classical digital system.

People seem to have this incorrect notion that quantum computers are just faster and better at everything.

Click to expand...

Very right. Saved me time.
Maybe in the future we'd have a Quantum Chip somewhere in the circuit board along with the other ICs.

@gelgoog indeed that maybe the case as future unfolds. But I think the industry has so much invested into current ICs that the inertia of it alone along with the magnitude of change that must take place will mean QCs will see a gradual introduction ( if they ever get out of labs and get commercialized).

It's been a year and 3 months or Sept 15 2020 to be exact when total sanction/ban on Huawei was imposed by the US. Both SMIC with its 14nm chips and TSMC owned 7nm and 5nm were forced to complied. Fast forward today it was reported by Long Xin (needed to ascertain his claimed) that SMIC will be supplying 14nm chips to Huawei. From what we know, an indigenous 14nm line will be achieved next year. Had SMIC found a way around the sanction or due to urgency the localized 14nm line had gone through its phase production?

An excerpt from a YT provider Long Xin ( will not post the entire video, for members that is interested just searched it in YT)

Nowadays, a lots media reports that SMIC may implement a 14nm chip manufacturing process to produce OLED screen driver chips for Huawei. Taking into account the well-known reasons before, this may mean that SMIC's 14nm process has achieved complete independence. Due to the relevant pressure imposed by foreign countries on Huawei, SMIC produces chips for Huawei's foundry, which is also a sign of China. The core 14nm chip manufacturing process has achieved independence and does not need to rely on the technical assistance of other countries to complete it. This is also a new breakthrough in the field of independent manufacturing and production of chips in China. As early as 2019, due to well-known reasons, TSMC, in the face of huge pressure, emphasized that its 7nm and more advanced 5nm processes used American technology in the proportion of less than 10%, so it can continue to produce for Huawei. Chips, SMIC uses the 14nm process to produce chips for Huawei.

Yukiolweipqled51 said:

Fair to say China will be playing catch up in the semiconductor industry way past 2030? @FairAndUnbiased @WTAN and others, your takes plz

Click to expand...

nope. in fact China has taken a leading position in many cases.

BYD Semiconductor Independently Develops 1200V Power Device Driver Chip BF1181​

BYD Semiconductor Co. ltd, a subsidiary of BYD, announced on Tuesday that . Based on previous R&D results the chip could begin shipping to major manufacturers in batches in December of this year.

海鸥300 Noise tests

Beginning on November 23, members of the research team of the Aviation Industry Flight Test Center came to Anyang one after another to carry out the airworthiness and noise test flight of the "Seagull" 300 aircraft. "Seagull" 300 is a multi-purpose light amphibious aircraft. Although the aircraft is light, the task of airworthiness and noise test flight is heavy, because the research team has never carried out related work on propeller aircraft.

The "Seagull" 300 aircraft airworthiness noise test flight team came from the original crew of the C919 aircraft airworthiness noise test flight. This team experienced the 2013 ARJ21-700 aircraft airworthiness noise test flight.

gadgetcool5 said:

China should focus on quantum computing, because it is less than 2 years behind the West at most. This area has far greater potential rewards for China than traditional EUV semiconductors, let alone DUV. It is far easier to get ahead in a new and emerging area than an already established area. By 2030 China could easily be ahead of the West in quantum computing.

If you look at 5G, the reason why Huawei was able to get ahead is because telecommunications technology was still in its infancy then. There was uncertainty in standards between CDMA and GSM, the US bet on CDMA and lost. Hence Huawei had an opening.

Similarly, currently the opening is in quantum computing. It's not that China should not invest in traditional semiconductors, but that even with massive investment it may still be playing catch up in 2030. You really want to be ahead, not playing catch up. The best quantum computers in the future will run laps around the best traditional semiconductors.

Click to expand...

Quantum computing is only useful at things you can not do practically in normal computers, such as large integer factoring. And China is not behind at all in that, and silicon semiconductors is not a dead end technology. I am in the industry. I can not remember how many times I heard silicon chip was dead. Silicon chips are extremely well rounded and well rooted. They will continue to be the default semiconductors for at least 20 years.

tokenanalyst said:

Huaxing Yuanchuang: Micro OLED testing equipment has been verified by Sony and end customers​

Please, Log in or Register to view URLs content!

According to the micro-net news, on December 14th, Huaxing Yuanchuang (688001.SH) stated on the investor interaction platform that Huaxing Yuanchuang Micro OLED has extremely high resolution, and its ultra-high resolution feature makes it the most suitable for AR. /VR equipment near-eye display technology products are expected to be widely used in near-eye display products in the future.

At the same time, it revealed that Huaxing Yuanchuang Micro OLED testing equipment has been verified by downstream customers Sony and end customers, and is currently cooperating with downstream customers to conduct preliminary research and development trials for subsequent product mass production, and its product competitiveness is at the forefront of the industry.

Please, Log in or Register to view URLs content!

Click to expand...

I wonder when the US will start demanding its allies not to buy semiconductors from China because if the US doesn't, its semiconductor industry will be dead. All because the US made a bad bet that when it denied China semiconductors, they believed China would surrender and submit to all US demands and not decide to make domestic semiconductors because they believed Chinese were too dumb to do it themselves.