Chinese semiconductor industry

Hendrik_2000 said:

I don't think this is true my god lithography is just a piece of technology created by man and not by god. If you look at China progress over the year, they did catch up in all of technology field except semiconductor. I can name aerospace not too long ago the western analyst mock Chinese airforce as antiquated air force yet within span of decade it can give run to the best of western airforce , Hight T/W turbo jet again we now see that China did produce their home grown jet engine comparable to the the best that the like of GE and Pratt and whitney. Naval design and Shipping in general again China is the largest shipyard in the world and churned out ship like dumpling. What remarkable is that they achieve it while under strictest tech embargo by the west

Land transport, China domestics car industry capturing larger and larger of domestic market where it was once dominated by the like of VW, Honda, GM and now even ahead in self driving car and EV
High speed train China now has the largest network of high speed train and has the most advance technology in HSR

I can go on and on Aerospace and Space and satellite industry again china now has their own space station and landed on Mars and Beidou GPS is up and running. Industry after industry China catch up to the west

Semiconductor is no different it just technology given time China will catch and even surpass the west It is just the latest technology that China has to conquer. I don't believe the first move advantage. At one time Intel is the most advance semiconductor company but they slip up and fall behind the Asian giant like TSMC
China was behind because the Chinese industry does not want to invest in FAB They naively believe they can depend on the western supplier and merrily making money without investing in manufacturing THEY ARE WRONG AND THEY NOW PAY HEFTY SUM.

But now everybody on board with having secure supply line and once China make up their mind they will reached their goal They prove it over the year in myriad field of technology and given that they produce million of STEM graduate combine with hardworking and immense research facility that they built over the year it just mater of time that they will also conquered SEMI. Mark my word

Click to expand...

I dream the day you can buy and custom your own 5nm SoC at Huaqiangbei.

Btw @FairAndUnbiased what's your opinion about the 02 test thing discussed a few pages back.

ericlfh said:

The


SMIC need to use domestic DUV as soon as they are ready, gradually phasing out ASML litho when domestic DUV matures. Huawei need to support domestc IC development and vice versa. Hope to buy my first Huawei phone next year.

Click to expand...

@ericlfh bro aside from Nokia, I had been using Huawei phone all my life , My traitorous son switch last year from Huawei to Apple and he gave me his P30 pro replacing my favorite Mate 20 pro and so far NO complained. I will wait when Hisilicon reintroduced its Kirin 980 7nm with 3D chiplet tech, that will take time maybe mid 2023 but at least the performance level will be less than a generation behind ( TSMC 3NM CHIP).

Regarding replacing ASML, bro it wouldn't happen with the massive investment being spend, it is foolish thing to do, it's a business venture need those ROI . What SMIC will do as @weig2000 had previously stated is the Dual approach strategy, BUT with future investment concentrated in procuring domestic equipment.

ansy1968 said:

@ericlfh bro aside from Nokia, I had been using Huawei phone all my life , My traitorous son switch last year from Huawei to Apple and he gave me his P30 pro replacing my favorite Mate 20 pro and so far NO complained. I will wait when Hisilicon reintroduced its Kirin 980 7nm with 3D chiplet tech, that will take time maybe mid 2023 but at least the performance level will be less than a generation behind ( TSMC 3NM CHIP).

Regarding replacing ASML, bro it wouldn't happen with the massive investment being spend, it is foolish thing to do, it's a business venture need those ROI . What SMIC will do as @weig2000 had previously stated is the Dual approach strategy, BUT with future investment concentrated in procuring domestic equipment.

Click to expand...

I used Xiaomi since their very first Xiaomi 1A which suck btw. I watch them grow from being a laughing stock to one of the top phones. Looks like I will switch to Huawei now .

Same with China ICs, I will use from now on and hope to see them grow from being a laughing stock (to some members here at least) to one of the world's best.

ericlfh said:

I used Xiaomi since their very first Xiaomi 1A which suck btw. I watch them grow from being a laughing stock to one of the top phones. Looks like I will switch to Huawei now .

Same with China ICs, I will use from now on and hope to see them grow from being a laughing stock (to some members here at least) to one of the world's best.

Click to expand...

@ericlfh bro you will like the tech it offer, I will tell you a funny story, while my son is teaching me about P30 PRO super zoom video and picture we had captured something amazing happening 2 blocks away from our balcony...lol and I had to delete it asap due to my son being underage...hahaha

Annihilation98 said:

This is how EUV looks like. This machine is gigantic and very sophisticated machine. I dont think China can build this even 10 years or more without espionage.
View attachment 79152

Click to expand...

That is physics not some magic that was revealed only to Caucasians by some oracle.
Any country from Somalia to Nepal to North Korea to Zimbabwe can build such though it would take long.
Look, the year is 2021 and China is not starting from a zero knowledge base and even though it took the West 20 years, it will take China just 5. Why?

Because catch up is always easier than pushing the frontier.

02 failure news came from 马督工, an independent online news/analysis show. His team consists mostly civil engineers and has no knowledge of semiconductor industry (I know because I have been watching his show on YouTube for years). His opinion is based on the fact there is no official news
This piece of shitty “news” doesn’t worth discussing

Annihilation98 said:

This is how EUV looks like. This machine is gigantic and very sophisticated machine. I dont think China can build this even 10 years or more without espionage.
View attachment 79152

Click to expand...

There are bigger machines that that... Your assertion is baseless given that China already has pre requisite technology capabilities in all manner of IC chip manufacturing processes.

caudaceus said:

I dream the day you can buy and custom your own 5nm SoC at Huaqiangbei.

Btw @FairAndUnbiased what's your opinion about the 02 test thing discussed a few pages back.

Click to expand...

I have no opinion about that because I cannot distinguish if that news is true or false, and what the importance of such a news is if it is true. What I do know is that and is equal to Canon in front end lithography since both are used for 90 nm processes.

Canon is an old school powerhouse in optical technologies, demonstrating that SMEE at the very least has been proven to do exactly what Canon has done. All I have to ask is - if they said Canon failed, who would believe them?

gadgetcool5 said:

The West was already doing lithography at the 28nm node ten years ago. If China invested tremendously in traditional lithography for the next ten years, where do you think they would be in 10 years from now? If you are optimistic, then they would have reached the 5nm, 3nm or 2nm nodes, roughly similar to where the West is today. But by then the West would have moved on with another ten years of development. Even the most optimistic people here say that China only needs "good enough" chips for most needs and don't claim China can surpass the West.

Click to expand...

No, this entire statement is factually false. Literally every word starting with "the west" is false.

What is the specific date of release of 28 nm chips?

What are the specific companies and fabs capable of 28 nm fabrication and how many are in 'the west'?

tokenanalyst said:

I understand and will be excellent, ideal to have a team of individuals that have done before exactly what you company wants to implement . That is the Ideal OPAMP, but reality is different, is difficult to poach someone who very especially knows what your company wants to solve and even dangerous because the company doesn't know if the poached person is going to implement a previous patented solution . What a company do is to form a team from people with people with knowledge closely related to the problem in question from Academia and other industries to tackle the company problem. The people working in the Chinese lithography problem come from Academia or other industries, i don't think none of them have worked in Nikon or ASML. The same goes for the people who develop Nikon first Immersion scanner or the people who worked in the first model of ASML EUV scanner. Maybe i am wrong about this or worse i am misinterpreting what you are saying, but that is the way i view the situation.

Click to expand...

basically the issue is that it is much more difficult to solve the edge case problems than the general problems. I'd say, and this is in very general terms, that there's a U distribution of difficulty for any enterprise that results in bimodal distribution of enterprise sizes ranging from small to near global monopoly.

1. Starting out is very hard. It is extremely hard to become a supplier in semiconductor equipment at all. Who will buy your stuff? You can only sell to one of a comparatively few fabs.

2. But once you have started out, it's relatively easy to grow and gain market share up to just before the leading edge since your competitors are far ahead and don't compete in your market segment, leaving a low cost niche to be exploited.

3. Once you are near the leading edge, it is hard again, because you need to invest more and more for incrementally smaller gains. This is the stage that Chinese semiconductor equipment companies are approaching.

Hiring people from academia is good. It provides a fresh perspective, expertise in research (for MS/PhD) and strong basic skills, important for 3 things: 1. cutting edge research when already at the leading edge, 2. sustaining engineering, 3. new ideas for startups.

But when you are approaching, but not yet at, the industry leading edge, what you really need is the state of the art in industry to even know what the issuse are.

For example, deposition. It's hard to make a deposition chamber. You need hundreds of thousands or a few millions of USD to make one. But actually, if you actually have the money to risk, you can make it from readily available commerical components like mass flow controllers, heated stages, vacuum chambers, etc. as an individual or small team. T. They have market values ~10's of millions USD. Then... you have nothing from 10's of millions to 1 billion USD size.

What separates them? Why can't Angstrom Engineering which makes deposition systems compete with AMAT or Lam Research in deposition systems?

An R&D scale system just has to work at all and not break down. It can be manually loaded and unloaded. It doesn't have to have say, 5G connectivity and interface with wafer handling robots. If it produces the correct film 99.99% of the time, it's good. It doesn't need inline metrology, I'm going to take the sample out and measure it on my own instruments.

A commercial production scale system doesn't need to just work and not break down. It has to interface with all the other equipment in the lab. It has to be autoloading and interface with robots. 99.9% correct is shit, that's not even ppm level precision.

OK, how do you go from a R&D scale to commercial scale tool? You can't learn this in a university. Nor should you - it's not leading edge stuff. It's mundane industry standards like "what is the standard gate valve used" and "what is the communications protocol of semiconductor equipment". It's stuff like "what are the trace metals to avoid" "what are industry cleanliness standards" and things like that. You need someone that already knows to handle stuff like this. Otherwise often times your product can just get rejected without explanation.