Chinese semiconductor industry

Status
Not open for further replies.
T

tacoburger

Junior Member
Registered Member
sunnymaxi said:
The money for equipment is being allocated to the fabs so the fabs ultimately make the decisions about which equipment providers and types of equipment get funding. basically indirectly money given to equipment makers. last year Chinese tools makers won record bid in mainland.
Click to expand...
That for equipment that's already developed and mass produced, and thus for companies that are already established. I'm talking about new companies and startups being founded. You know, a group of guys starting up a brand new business to make lithography systems and whatnot. Because I'm seeing dozens of brand new chip design companies but not a lot of new equipment/materials companies and not a single new company that's focused on lithography systems.
sunnymaxi said:
you should stop behave like a child. how many times our experts cleared your doubt.

SMEE have successfully developed 28nm DUVi .. currently under validating process. if everything goes well then small scale production can start at the end of this year.
Click to expand...
It's also pretty clear that it's way way overschedule and probably overbudget. They were supposed to have it done and in mass production by 2021 or 2022. Even then, we have no idea how long it's gonna take to actually to reach mass production, enough to supply China's domestic needs. This is at a critical time in the industry, where sanctions are being dropped left and right and ASML can't 100 guarantee supply.

This isn't some random company, it's probably the most important crucial part of the entire semiconductor supply that China has the least control over. We don't have details, but they're probably swimming in money and resources, they are more or less the only ones working on DUV right now. When things are this high stakes and important, and you have the money and resources, you don't get to excuse incompetence. Can you imagine if the development of the nuclear bomb took twice as long as it did? It's not like any of their previous projects inspire confidence in the company.
 
sunnymaxi

sunnymaxi

Major
Registered Member
tacoburger said:
That for equipment that's already developed and mass produced, and thus for companies that are already established. I'm talking about new companies and startups being founded. You know, a group of guys starting up a brand new business to make lithography systems and whatnot. Because I'm seeing dozens of brand new chip design companies but not a lot of equipment/materials companies and not a single new company that's focused on lithography systems.
Click to expand...
there are a lot material/equipment related semiconductor companies being founded in past decade. please check thoroughly on this thread. AMEC/NAURA covered almost all basic tools. and few more like kingsemi and others. i can't post every name. @tokenanalyst post chart of complete supply chain of semiconductor tools/materials. Chinese firms have covered all tools for 28nm and some are on 14nm. slowly slowly they are climbing up the value chain.

do you know, Beijing stock exchange and local government programs have especially established for semiconductors. each region/province providing funds to those little giants. SMEs.

apart from SMEE, one Beijing based firm ''CETC'' also working on lithography

now come to EUV. entire Chinese nation currently working on EUV include Huawei. many Chinese institutes deeply involved in subsystems development. so far we have seen 8 to 9 EUV subsystem patents. posted by @tokenanalyst

tacoburger said:
not a single new company that's focused on lithography systems.
Click to expand...
LOL. this is funny bro.

SMEE is just integrating all subsystems and build a prototype. all subsystem developed by local Chinese firms/institutes. mostly are unknown. there are many unknown heroes to work on lithography subsystems.

they are building new production facility for light source

tacoburger said:
It's also pretty clear that it's way way overschedule and probably overbudget. They were supposed to have it done and in mass production by 2021 or 2022. Even then, we have no idea how long it's gonna take to actually to reach mass production, enough to supply China's domestic needs. This is at a critical time in the industry, where sanctions are being dropped left and right and ASML can't 100 guarantee supply.
Click to expand...
is DUVi that easy to develop? LOOL

do you know why SMEE has take this much time to produce DUVi prototype ? coz domestically produced components were delayed. and as of now, all major subsystem and components of SMEE's DUVi are indigenous. this is the actual progress. and it will take some time to start mass production.

SSA800i currently under validating process. 5 units can be produced in this year. if everything goes well. you see they will start small scale production.
 
L

latenlazy

Brigadier
tacoburger said:
That for equipment that's already developed and mass produced, and thus for companies that are already established. I'm talking about new companies and startups being founded. You know, a group of guys starting up a brand new business to make lithography systems and whatnot. Because I'm seeing dozens of brand new chip design companies but not a lot of new equipment/materials companies and not a single new company that's focused on lithography systems.

It's also pretty clear that it's way way overschedule and probably overbudget. They were supposed to have it done and in mass production by 2021 or 2022. Even then, we have no idea how long it's gonna take to actually to reach mass production, enough to supply China's domestic needs. This is at a critical time in the industry, where sanctions are being dropped left and right and ASML can't 100 guarantee supply.

This isn't some random company, it's probably the most important crucial part of the entire semiconductor supply that China has the least control over. We don't have details, but they're probably swimming in money and resources, they are more or less the only ones working on DUV right now. When things are this high stakes and important, and you have the money and resources, you don't get to excuse incompetence. Can you imagine if the development of the nuclear bomb took twice as long as it did? It's not like any of their previous projects inspire confidence in the company.
Click to expand...
Buddy you need to distinguish between funding for mass deployment vs funding for tech development. Not the same things. The most expensive part of industrial development is not developing the technology but scaling production. They’ve already developed most of the technology they need in the equipment space. They need to scale production now or all that technology development won’t matter. Having customers decide what gets purchased is how you scale production efficiently.

But also guess what a good share of the profits from mass deployment ends up getting used for :rolleyes:
 
tphuang

tphuang

Lieutenant General
Staff member
Super Moderator
VIP Professional
Registered Member
tphuang said:
Primer on the 3 fabs formed by CEC, Shenzhen gov't & Huawei in Shenzhen that will start production over the next 2 years

1 for advanced process, 1 for memory chips and 1 for mature process.

Given CXMT's struggles to reach 17nm, I think Huawei might produce the most advanced DRAMs in China in a few years.
Click to expand...
b0a79ab3gy1hcqagpw83qj20qc192dvr.jpg
Alright so on top of what was said yesterday, this is apparently a Hisilicon employee posting this news from Shanghai. I think if true, this confirms the theory that SMSC has been producing 7nm chips for Hisilicon since the new PXW fab has yet to start production. Also makes sense for Hisilicon to work with SMSC fab which is next to it.

He said Ascend chips are resurrected. After that, Tiangang base station chip will be completed.
Followed by Balong 5G chipset
Finally Kirin smartphone SoC
 
FairAndUnbiased

FairAndUnbiased

Brigadier
Registered Member
tacoburger said:
Why is there so much money going into chip design and dozens of chip design companies and so little going to equipment or other necessary stuff like photoresist, optical systems and masks? Really seems like for every equipment manufacturer, there's like 5 semiconductor design companies. Seems like a big waste, do you really need so many design companies?

If you want have reductant companies, why not have a competitor to SMEE? Seems insane that such a vital part of the supply chain is in the hands of a underperforming company. I feel like the size of china's semiconductor industry could easily have allowed for two or even three companies dedicated to lithography systems, maybe even private ones. It's not like the need for DUV machines will go away even if you get access to EUV. They will always be used for mature chips.
Click to expand...
yes you do.

Every market is like a pyramid. The closer you are to the base, the customer, the more money there is. Entire semiconductor wafer market is 10 billion USD. ASML dominates the lithography market, near monopoly in fact. Yet
Please, Log in or Register to view URLs content!


Compare:
Please, Log in or Register to view URLs content!
.
Please, Log in or Register to view URLs content!
That is to say, automotive chips alone built using refurbished 1990's equipment, pays enough for 3 ASMLs.

This is because the ultimate source of all money in all physical product markets is the final customer. The more B2B and upstream your product is, the smaller the market, because at every step away from the customer, a company must make profit to stay afloat. They can make less profit, but they can't make 0 profit, otherwise they can't pay their workers and suppliers. So the market size of their customers must be larger than their own collective market size.

Chip designers are very close to the final customer: they are usually just 1-2 step away from civilian customers.

But how far is ASML from the final customer?

Customer -> Industrial/consumer products -> Electronics OEM (module, packaging, PCBs, etc) -> Semiconductor designer -> Semiconductor fab -> Semiconductor Equipment -> Semiconductor Equipment Components/Materials

That is why if you want semiconductor independence, you need to invest in the low levels first. It is way better to have 10 fabs that can all do 90-180 nm with KrF or dry ArF and at least be able to make TVs, cars, appliances, etc. than to have 1 EUV 3 nm fab that can make bleeding edge server chips and GPUs but then there isn't enough capacity for TV, car, appliance chips.

Then your TV, car and appliance companies import chips because you physically can't produce enough chips for them.

That is why you need tons of semiconductor designers to buy fab services and produce niche designs for customers, otherwise there isn't enough demand for more fabs, and the fabs either won't scale up or they scale up then sit idle and lose money. Fabs can't just be turned off easily, so they will spend money regardless.
 
L

latenlazy

Brigadier
FairAndUnbiased said:
yes you do.

Every market is like a pyramid. The closer you are to the base, the customer, the more money there is. Entire semiconductor wafer market is 10 billion USD. ASML dominates the lithography market, near monopoly in fact. Yet
Please, Log in or Register to view URLs content!


Compare:
Please, Log in or Register to view URLs content!
.
Please, Log in or Register to view URLs content!
That is to say, automotive chips alone built using refurbished 1990's equipment, pays enough for 3 ASMLs.

This is because the ultimate source of all money in all physical product markets is the final customer. The more B2B and upstream your product is, the smaller the market, because at every step away from the customer, a company must make profit to stay afloat. They can make less profit, but they can't make 0 profit, otherwise they can't pay their workers and suppliers. So the market size of their customers must be larger than their own collective market size.

Chip designers are very close to the final customer: they are usually just 1-2 step away from civilian customers.

But how far is ASML from the final customer?

Customer -> Industrial/consumer products -> Electronics OEM (module, packaging, PCBs, etc) -> Semiconductor designer -> Semiconductor fab -> Semiconductor Equipment -> Semiconductor Equipment Components/Materials

That is why if you want semiconductor independence, you need to invest in the low levels first. It is way better to have 10 fabs that can all do 90-180 nm with KrF or dry ArF and at least be able to make TVs, cars, appliances, etc. than to have 1 EUV 3 nm fab that can make bleeding edge server chips and GPUs but then there isn't enough capacity for TV, car, appliance chips.

Then your TV, car and appliance companies import chips because you physically can't produce enough chips for them.

That is why you need tons of semiconductor designers to buy fab services and produce niche designs for customers, otherwise there isn't enough demand for more fabs, and the fabs either won't scale up or they scale up then sit idle and lose money. Fabs can't just be turned off easily, so they will spend money regardless.
Click to expand...
China learned this lesson the hard way actually. In the 90s they tried to build a number of strategic industries from upstream production to downstream end goods first rather than from downstream end goods to upstream production, and each one was riddled with inefficient unprofitable failures. The strategy that ended up working for China was letting the state support technology development but letting the market support industry development. We are exiting the state supported technology development phase of the China’s semiconductors industry and entering the market supported industry development phase.
 
C

CMP

Senior Member
Registered Member
FairAndUnbiased said:
yes you do.

Every market is like a pyramid. The closer you are to the base, the customer, the more money there is. Entire semiconductor wafer market is 10 billion USD. ASML dominates the lithography market, near monopoly in fact. Yet
Please, Log in or Register to view URLs content!


Compare:
Please, Log in or Register to view URLs content!
.
Please, Log in or Register to view URLs content!
That is to say, automotive chips alone built using refurbished 1990's equipment, pays enough for 3 ASMLs.

This is because the ultimate source of all money in all physical product markets is the final customer. The more B2B and upstream your product is, the smaller the market, because at every step away from the customer, a company must make profit to stay afloat. They can make less profit, but they can't make 0 profit, otherwise they can't pay their workers and suppliers. So the market size of their customers must be larger than their own collective market size.

Chip designers are very close to the final customer: they are usually just 1-2 step away from civilian customers.

But how far is ASML from the final customer?

Customer -> Industrial/consumer products -> Electronics OEM (module, packaging, PCBs, etc) -> Semiconductor designer -> Semiconductor fab -> Semiconductor Equipment -> Semiconductor Equipment Components/Materials

That is why if you want semiconductor independence, you need to invest in the low levels first. It is way better to have 10 fabs that can all do 90-180 nm with KrF or dry ArF and at least be able to make TVs, cars, appliances, etc. than to have 1 EUV 3 nm fab that can make bleeding edge server chips and GPUs but then there isn't enough capacity for TV, car, appliance chips.

Then your TV, car and appliance companies import chips because you physically can't produce enough chips for them.

That is why you need tons of semiconductor designers to buy fab services and produce niche designs for customers, otherwise there isn't enough demand for more fabs, and the fabs either won't scale up or they scale up then sit idle and lose money. Fabs can't just be turned off easily, so they will spend money regardless.
Click to expand...
Sounds like a strong argument for full vertical integration.
 
T

tokenanalyst

Brigadier
Registered Member

Semiconductor precision parts business is growing rapidly Jiangfeng Electronics' net profit last year increased by 148.72% year-on-year.​


Jiwei.com reported on April 7 that Jiangfeng Electronics disclosed its annual report. In 2022, the company will achieve operating income of 2.324 billion yuan, a year-on-year increase of 45.80%; billion, a year-on-year increase of 186.50%.

In terms of products, ultra-high-purity targets achieved sales revenue of 1.611 billion yuan, a year-on-year increase of 36.40%. It pointed out that the company actively improves the sales service network and guarantee base, serves customers nearby, continuously tracks customer needs, provides on-site PM, product surveying and mapping, reverse development docking and other services, closely interacts with customers, wins customers' trust, strengthens market penetration, Expand market share and increase sales scale steadily.

The parts business achieved sales of 358 million yuan, a year-on-year increase of 94.51%. According to Jiangfeng Electronics, as global wafer manufacturers invest in new production capacity to varying degrees, which in turn drives the development of the semiconductor precision parts industry, semiconductor precision parts have broad market prospects. The semiconductor precision parts business is closely related to the semiconductor target business. The company has successfully applied the long-term accumulated technical capabilities, quality assurance capabilities, and customer understanding capabilities in the field of ultra-high-purity metal sputtering targets for semiconductors to the field of semiconductor precision parts. Realized the rapid growth of semiconductor precision parts business.

At the same time, the parts and components produced by Jiangfeng Electronics include equipment manufacturing parts and process consumable parts, which can be widely used in PVD (Physical Vapor Deposition), CVD (Chemical Vapor Deposition), etching machines, CMP (Chemical Mechanical Planarization), etc. In semiconductor equipment, its production process requires extremely high technical requirements such as metal material precision manufacturing technology and special surface treatment technology. The products are not only sold to wafer manufacturers as equipment consumables for periodic replacement, but also sold to semiconductor equipment manufacturers. in equipment production. In addition, the company entered the field of ceramic copper-clad substrates in a timely manner. At present, it has built the first domestic production line with the world's advanced level and independent design of the third-generation semiconductor power device module core material manufacturing line. The planning and construction have independent intellectual property rights and advanced technology. , a large-scale production base of localized copper-clad ceramic substrates with complete material specifications and automated production lines.

Please, Log in or Register to view URLs content!
 
FairAndUnbiased

FairAndUnbiased

Brigadier
Registered Member
tokenanalyst said:

Semiconductor precision parts business is growing rapidly Jiangfeng Electronics' net profit last year increased by 148.72% year-on-year.​


Jiwei.com reported on April 7 that Jiangfeng Electronics disclosed its annual report. In 2022, the company will achieve operating income of 2.324 billion yuan, a year-on-year increase of 45.80%; billion, a year-on-year increase of 186.50%.

In terms of products, ultra-high-purity targets achieved sales revenue of 1.611 billion yuan, a year-on-year increase of 36.40%. It pointed out that the company actively improves the sales service network and guarantee base, serves customers nearby, continuously tracks customer needs, provides on-site PM, product surveying and mapping, reverse development docking and other services, closely interacts with customers, wins customers' trust, strengthens market penetration, Expand market share and increase sales scale steadily.

The parts business achieved sales of 358 million yuan, a year-on-year increase of 94.51%. According to Jiangfeng Electronics, as global wafer manufacturers invest in new production capacity to varying degrees, which in turn drives the development of the semiconductor precision parts industry, semiconductor precision parts have broad market prospects. The semiconductor precision parts business is closely related to the semiconductor target business. The company has successfully applied the long-term accumulated technical capabilities, quality assurance capabilities, and customer understanding capabilities in the field of ultra-high-purity metal sputtering targets for semiconductors to the field of semiconductor precision parts. Realized the rapid growth of semiconductor precision parts business.

At the same time, the parts and components produced by Jiangfeng Electronics include equipment manufacturing parts and process consumable parts, which can be widely used in PVD (Physical Vapor Deposition), CVD (Chemical Vapor Deposition), etching machines, CMP (Chemical Mechanical Planarization), etc. In semiconductor equipment, its production process requires extremely high technical requirements such as metal material precision manufacturing technology and special surface treatment technology. The products are not only sold to wafer manufacturers as equipment consumables for periodic replacement, but also sold to semiconductor equipment manufacturers. in equipment production. In addition, the company entered the field of ceramic copper-clad substrates in a timely manner. At present, it has built the first domestic production line with the world's advanced level and independent design of the third-generation semiconductor power device module core material manufacturing line. The planning and construction have independent intellectual property rights and advanced technology. , a large-scale production base of localized copper-clad ceramic substrates with complete material specifications and automated production lines.

Please, Log in or Register to view URLs content!
Click to expand...
this is for things like ceramic coated aluminum chambers, pure silicon mechanical components, etc. this represents very high level of materials science tech.
 
Status
Not open for further replies.
Top