Chinese semiconductor industry

european_guy said:

Also Reuters reports this:

Please, Log in or Register to view URLs content!

The source is

Please, Log in or Register to view URLs content!

, a Chinese official financial newspaper with headquarters in Shenzhen (Huawei's city).

Reuters clarifies that the expected 40M sales are just for 2023:



This is beyond impressive to me! How much wpm capacity is needed for 40M phones? Assume they started producing in March, it is 8/9 months production for 40M phones, it means a yearly capacity of 60M Kirin 9000s chips. That's huge!

Mainly they have fully counterbalanced any possible ban on Nvidia AI: at this point a US ban would give the whole Chinese AI market to Huawei....even the most hawkish US think tanker cannot be that stupid!


.....or OTH it is maybe possible China is bluffing on it, a big, royal bluff just to avoid possible additional bans, and to gain precocious time while US remains at the window to see if SMIC 7nm capacity is for real.

I would not think is a bluff, but I'm very curious to see how this thing develops.

Click to expand...

so the 40m, most of it is not with Kirin 9000S, since they probably achieved like 15m in first half of the year. And even for second half, good chunk of it are with older models that don't use kirin chips.

I think if they get to 10m this year with Kirin chips, it will be pretty good (just because of the late launches)

On the other hand, next year actually will be mostly Kirin chips, so they need to get ready for a lot of production if they want to sell 60m phones.

Sichuan Zigong signed a 1 billion yuan high-end photoresist and ultra-clean semiconductor functional chemicals project​

A domestic technology company signed a project cooperation agreement with Yantan District, Zigong City, Sichuan, planning to invest in an annual output of 100,000 tons of high-end photoresist and ultra-clean semiconductor functions Chemical industrialization project.
According to news released by Zigong, the project to be settled in Zigong City plans to invest 1 billion yuan and covers an area of about 140 acres. It plans to build photoresist, electronic grade hydrogen peroxide, electronic grade nitric acid, electronic grade hydrofluoric acid, etching solution, developer, Production lines for stripping fluid, cleaning fluid, high-purity electronic grade sulfuric acid and solvent recovery.
It is reported that after the project is completed and reaches production capacity, it is expected that the annual sales revenue will exceed 1.7 billion yuan, the annual tax revenue will be 300 million yuan, and it will provide more than 100 jobs.

Chinatalk Jordan Schneider finally opens his mouth. Although he is better than many western outlets out there that still rely on yields as the weapon to attack Huawei Breakthrough. He attacked with China is stuck at 7NM and without western semiconductor equipment, wont be able to progress further for decades.

So, my question for the people who follow the semi-conductor industry, how true is this. If US stops selling of all semiconductor equipment to China, how long would it take for China to develop everything locally?

With my limited knowledge of Semiconductor tech, my impression is that semiconductor tech is not that hard. Its just as cutting edge as compared to say precision engineering required for making high quality jet engines. The only difference is Semiconductor tech was never weaponised and thus powerful countries never needed to worry about not having it.

Thus they allowed full market principles to work here. Which means they imported semiconductors as much as they could from the cheapest seller. Unlike say jen engines, which was always restricted technology and China had to invest on it for decades and are now close to the cutting edge of US tech.

Now that semiconductor has been weaponised, a global supply chain is simply not acceptable. Everything has to be localized and it will be.

But the question to people here, how long will that take? 5 years, 10 years? or just 2 years as some rumours suggest China will use its own EUV in 2 years.

tamsen_ikard said:

Please, Log in or Register to view URLs content!

Chinatalk Jordan Schneider finally opens his mouth. Although he is better than many western outlets out there that still rely on yields as the weapon to attack Huawei Breakthrough. He attacked with China is stuck at 7NM and without western semiconductor equipment, wont be able to progress further for decades.

So, my question for the people who follow the semi-conductor industry, how true is this. If US stops selling of all semiconductor equipment to China, how long would it take for China to develop everything locally?

With my limited knowledge of Semiconductor tech, my impression is that semiconductor tech is not that hard. Its just as cutting edge as compared to say precision engineering required for making high quality jet engines. The only difference is Semiconductor tech was never weaponised and thus powerful countries never needed to worry about not having it.

Thus they allowed full market principles to work here. Which means they imported semiconductors as much as they could from the cheapest seller. Unlike say jen engines, which was always restricted technology and China had to invest on it for decades and are now close to the cutting edge of US tech.

Now that semiconductor has been weaponised, a global supply chain is simply not acceptable. Everything has to be localized and it will be.

But the question to people here, how long will that take? 5 years, 10 years? or just 2 years as some rumours suggest China will use its own EUV in 2 years.

Click to expand...

These people never learn, they only know how to move the goalpost when they've been smacked in the face.

tamsen_ikard said:

Please, Log in or Register to view URLs content!

Chinatalk Jordan Schneider finally opens his mouth. Although he is better than many western outlets out there that still rely on yields as the weapon to attack Huawei Breakthrough. He attacked with China is stuck at 7NM and without western semiconductor equipment, wont be able to progress further for decades.

So, my question for the people who follow the semi-conductor industry, how true is this. If US stops selling of all semiconductor equipment to China, how long would it take for China to develop everything locally?

With my limited knowledge of Semiconductor tech, my impression is that semiconductor tech is not that hard. Its just as cutting edge as compared to say precision engineering required for making high quality jet engines. The only difference is Semiconductor tech was never weaponised and thus powerful countries never needed to worry about not having it.

Thus they allowed full market principles to work here. Which means they imported semiconductors as much as they could from the cheapest seller. Unlike say jen engines, which was always restricted technology and China had to invest on it for decades and are now close to the cutting edge of US tech.

Now that semiconductor has been weaponised, a global supply chain is simply not acceptable. Everything has to be localized and it will be.

But the question to people here, how long will that take? 5 years, 10 years? or just 2 years as some rumours suggest China will use its own EUV in 2 years.

Click to expand...

I would expect china can use foreign equipments immersion duv to make chips at 5nm in next 2 years, and then domestic euv will be used in 2026.

tamsen_ikard said:

Please, Log in or Register to view URLs content!

Chinatalk Jordan Schneider finally opens his mouth. Although he is better than many western outlets out there that still rely on yields as the weapon to attack Huawei Breakthrough. He attacked with China is stuck at 7NM and without western semiconductor equipment, wont be able to progress further for decades.

So, my question for the people who follow the semi-conductor industry, how true is this. If US stops selling of all semiconductor equipment to China, how long would it take for China to develop everything locally?

With my limited knowledge of Semiconductor tech, my impression is that semiconductor tech is not that hard. Its just as cutting edge as compared to say precision engineering required for making high quality jet engines. The only difference is Semiconductor tech was never weaponised and thus powerful countries never needed to worry about not having it.

Thus they allowed full market principles to work here. Which means they imported semiconductors as much as they could from the cheapest seller. Unlike say jen engines, which was always restricted technology and China had to invest on it for decades and are now close to the cutting edge of US tech.

Now that semiconductor has been weaponised, a global supply chain is simply not acceptable. Everything has to be localized and it will be.

But the question to people here, how long will that take? 5 years, 10 years? or just 2 years as some rumours suggest China will use its own EUV in 2 years.

Click to expand...

First of all, he doesn't have any evidence for the 50% yield claim. That number is by blob quoting other blob who quote anonymous sources. TechInsights actually argued for the opposite, after looking at features on the chip. Those people are actual experts who examined thousands of different chips. They know how a deficient process looks like. I'll take their word.

Furthermore, how it is relevant if China is stuck at 7 nm for a few years? SMIC N+2 is within 2 nodes of the leading edge process. In an era when the difference between two consecutive nodes is relatively small and getting smaller... It is more than enough for important stuff. Even commercially, it has a huge market, particularly because of the sanctions. Journalists care too much about names and "generations". SMIC N+7 and HiSilicon's design expertise has practical implications, of which none are good for US interests.

BoraTas said:

First of all, he doesn't have any evidence for the 50% yield claim. That number is by blob quoting other blob who quote anonymous sources. TechInsights actually argued for the opposite, after looking at features on the chip. Those people are actual experts who examined thousands of different chips. They know how a deficient process looks like. I'll take their word.

Furthermore, how it is relevant if China is stuck at 7 nm for a few years? SMIC N+2 is within 2 nodes of the leading edge process. In an era when the difference between two consecutive nodes is relatively small and getting smaller... It is more than enough for important stuff. Even commercially, it has a huge market, particularly because of the sanctions. Journalists care too much about names and "generations". SMIC N+7 and HiSilicon's design expertise has practical implications, of which none are good for US interests.

Click to expand...

Oh no...

TBH I always expected the US to tighten sanctions after this news came out. It’s way too embarrassing to announce less than 10 months ago this was happening, and then twist the arms of other countries to play along against their own economic interest … and just to give up because Huawei shipped one phone?

No way. The US will probably lock down all direct US products to SMIC / Huawei within the next month, no more Qualcomm licenses. They’ll try to lock down ASML, but they’ll fail.

cctang said:

TBH I always expected the US to tighten sanctions after this news came out. It’s way too embarrassing to announce less than 10 months ago this was happening, and then twist the arms of other countries to play along against their own economic interest … and just to give up because Huawei shipped one phone?

No way. The US will probably lock down all direct US products to SMIC / Huawei within the next month, no more Qualcomm licenses. They’ll try to lock down ASML, but they’ll fail.

Click to expand...

I think the value of the SMIC breakthrough isn't really to convince the Americans to relax the sanctions, but to dissuade the partners from following Washington's demands. It took the Biden admin a lot of effort to get the Dutch onboard, and this is what they've got to show for the suffering?

siegecrossbow said:

I think the value of the SMIC breakthrough isn't really to convince the Americans to relax the sanctions, but to dissuade the partners from following Washington's demands. It took the Biden admin a lot of effort to get the Dutch onboard, and this is what they've got to show for the suffering?

Click to expand...

Sir you have to include TSMC, there are a lot regrets and blaming going around as they help their competitor become a monster.