Chinese semiconductor industry

Status
Not open for further replies.
W

WTAN

Junior Member
Registered Member
Weaasel said:
I recently heard of nanoimprint lithography, which Canon has developed the equipment for, and which does not utilize projection lenses to make imprints of circuit patterns of a reticle unto photoresists resin layered unto semiconductor wafers, but instead "perfectly" positions the photoresist resin atop a wafer using inkjet technology and then also "perfectly" places and presses the reticle atop the photoresist and then upon the incidence of DUV light sources through the transparent portions of the reticle that impact the photoresists, circuit patterns capable of matching what can be achieved by ASML' s EUV machines (>10 nm) can be achieved.

Has China attempted any research and development into nanoimprint lithography? Does Canon have a good chance of commercially marketing its technology?
Click to expand...
Canon abandoned its DUVL manufacturing business and chose to go with nanoimprint Lithography.
Canon claims it will eventually be as good as DUVL and even EUVL with a much reduced cost in terms of equipment.
They have alot of experience with making Printers using Inkjet technology.
So this is something that they can apply inkjet tech to.
They are supplying their equipment to a new FAB being built in Japan. We will have to see how this works out.
There are i believe 2 companies in China developing nanoimprint Lithography.
So China is also pursuing this technology.
 
W

wxw456

New Member
Registered Member
ZeEa5KPul said:
If Japan has this technology, why does it utterly fail in getting any market share in EUV tools?
Click to expand...
To really understand Gigaphoton's predicament you have to travel back in time to the year 2000. In 2000 Nikon was the largest lithography machine supplier with two other major competitors ASML and Canon. At the turn of the millenium lithography machine companies were looking to develop Next-Generation Lithography (NGL) technology. Nikon was focused on the development of its electron beam projection (EPL) lithography tool and opted not to develop an extreme ultraviolet (EUV) lithography tool. Whereas ASML and Canon opted to develop an extreme ultraviolet (EUV) lithography tool.

Here is an EETimes article from 2001:
Please, Log in or Register to view URLs content!

Here is an article from 2001 about Canon:
Please, Log in or Register to view URLs content!

This was essentially a race/competition between EPL and EUV lithography for the next-generation lithography technology that started in 2000. 20 years later we know the outcome of this competition.

By 2008 Nikon, Canon and ASML decided that EUV lithography was the winner:
Please, Log in or Register to view URLs content!

Nikon's failed bet put them behind ASML in the EUVL development race:
Please, Log in or Register to view URLs content!

The problem was that Nikon's bet on EPL lost to EUV and that put the company a decade behind ASML. This leaves just Canon and ASML that started EUVL research right from 2000. The answer to why ASML beat out Canon in the race to EUVL is economic. In 2000, Nikon had 40% of the lithography-tool market share, ASML had about 30% market share, and Canon held 20% market share. ASML had more money and invested more into EUVL. When ASML won the race to develop the first commercial EUVL machine Nikon and Canon instantly lost billions of dollars in future sales and the first mover advantage. Canon eventually cancelled their EUVL development work. In 2008 the Canon SFET tool was still only for R&D. Whereas ASML in 2009 was planning to ship its "pre-production" EUV tool to customer sites.

In 2021 Gigaphoton may have a 250W EUV power source, but it doesn't have a customer. The only EUVL producer ASML has Cymer to supply its power source. Nikon and Canon dropped out of the EUVL market so Gigaphoton has no domestic customers. Only companies in China are actively working on developing a EUVL machine, but Gigaphoton can't/won't sell the product to them. So Gigaphoton is a company with a wonderful product without any customers (polite way to say business failure).
 
A

ansy1968

Brigadier
Registered Member
wxw456 said:
To really understand Gigaphoton's predicament you have to travel back in time to the year 2000. In 2000 Nikon was the largest lithography machine supplier with two other major competitors ASML and Canon. At the turn of the millenium lithography machine companies were looking to develop Next-Generation Lithography (NGL) technology. Nikon was focused on the development of its electron beam projection (EPL) lithography tool and opted not to develop an extreme ultraviolet (EUV) lithography tool. Whereas ASML and Canon opted to develop an extreme ultraviolet (EUV) lithography tool.

Here is an EETimes article from 2001:
Please, Log in or Register to view URLs content!

Here is an article from 2001 about Canon:
Please, Log in or Register to view URLs content!

This was essentially a race/competition between EPL and EUV lithography for the next-generation lithography technology that started in 2000. 20 years later we know the outcome of this competition.

By 2008 Nikon, Canon and ASML decided that EUV lithography was the winner:
Please, Log in or Register to view URLs content!

Nikon's failed bet put them behind ASML in the EUVL development race:
Please, Log in or Register to view URLs content!

The problem was that Nikon's bet on EPL lost to EUV and that put the company a decade behind ASML. This leaves just Canon and ASML that started EUVL research right from 2000. The answer to why ASML beat out Canon in the race to EUVL is economic. In 2000, Nikon had 40% of the lithography-tool market share, ASML had about 30% market share, and Canon held 20% market share. ASML had more money and invested more into EUVL. When ASML won the race to develop the first commercial EUVL machine Nikon and Canon instantly lost billions of dollars in future sales and the first mover advantage. Canon eventually cancelled their EUVL development work. In 2008 the Canon SFET tool was still only for R&D. Whereas ASML in 2009 was planning to ship its "pre-production" EUV tool to customer sites.

In 2021 Gigaphoton may have a 250W EUV power source, but it doesn't have a customer. The only EUVL producer ASML has Cymer to supply its power source. Nikon and Canon dropped out of the EUVL market so Gigaphoton has no domestic customers. Only companies in China are actively working on developing a EUVL machine, but Gigaphoton can't/won't sell the product to them. So Gigaphoton is a company with a wonderful product without any customers (polite way to say business failure).
Click to expand...
@wxw456 bro DARPA had a role they funded SEMATECH the consortium to challenge the Japanese dominance. Fast forward today the US and its allies try to recreate the same condition but there is a major problem, the Chinese are not in the dominant position they are and they are competing among themselves.
 
L

latenlazy

Brigadier
wxw456 said:
To really understand Gigaphoton's predicament you have to travel back in time to the year 2000. In 2000 Nikon was the largest lithography machine supplier with two other major competitors ASML and Canon. At the turn of the millenium lithography machine companies were looking to develop Next-Generation Lithography (NGL) technology. Nikon was focused on the development of its electron beam projection (EPL) lithography tool and opted not to develop an extreme ultraviolet (EUV) lithography tool. Whereas ASML and Canon opted to develop an extreme ultraviolet (EUV) lithography tool.

Here is an EETimes article from 2001:
Please, Log in or Register to view URLs content!

Here is an article from 2001 about Canon:
Please, Log in or Register to view URLs content!

This was essentially a race/competition between EPL and EUV lithography for the next-generation lithography technology that started in 2000. 20 years later we know the outcome of this competition.

By 2008 Nikon, Canon and ASML decided that EUV lithography was the winner:
Please, Log in or Register to view URLs content!

Nikon's failed bet put them behind ASML in the EUVL development race:
Please, Log in or Register to view URLs content!

The problem was that Nikon's bet on EPL lost to EUV and that put the company a decade behind ASML. This leaves just Canon and ASML that started EUVL research right from 2000. The answer to why ASML beat out Canon in the race to EUVL is economic. In 2000, Nikon had 40% of the lithography-tool market share, ASML had about 30% market share, and Canon held 20% market share. ASML had more money and invested more into EUVL. When ASML won the race to develop the first commercial EUVL machine Nikon and Canon instantly lost billions of dollars in future sales and the first mover advantage. Canon eventually cancelled their EUVL development work. In 2008 the Canon SFET tool was still only for R&D. Whereas ASML in 2009 was planning to ship its "pre-production" EUV tool to customer sites.

In 2021 Gigaphoton may have a 250W EUV power source, but it doesn't have a customer. The only EUVL producer ASML has Cymer to supply its power source. Nikon and Canon dropped out of the EUVL market so Gigaphoton has no domestic customers. Only companies in China are actively working on developing a EUVL machine, but Gigaphoton can't/won't sell the product to them. So Gigaphoton is a company with a wonderful product without any customers (polite way to say business failure).
Click to expand...
Worth also noting that when ASML’s first EUV instruments came out they found that its photon output wasn’t sufficient to meet the industry’s unit economy demands, which is another reason why EUV deployment into the fabrication process was delayed for so many years. A major cause was that initial laser driver for the LPP light source was much weaker, and increasing the power of the laser driver became a very messy physics problem of how to excite tin droplets at higher power without causing the droplet to scatter and spray. ASML spent almost another decade with Cymer to work on increasing the output of the light source sufficient enough to meet the wafer output speed needed to meet the high cost of the whole fabrication process. So when we discuss EUV development, keep in mind that being able have a light source that can already output similar power or higher than ASML’s EUV instrument is a very big deal for the speed of the development timeline. I don’t think it’s exaggerating to say that development speed will hinge predominantly on the light source problem.
 
W

wxw456

New Member
Registered Member
ansy1968 said:
@wxw456 bro DARPA had a role they funded SEMATECH the consortium to challenge the Japanese dominance. Fast forward today the US and its allies try to recreate the same condition but there is a major problem, the Chinese are not in the dominant position they are and they are competing among themselves.
Click to expand...
The problem for chip and equipment manufacturers in SK, Taiwan and Japan is that they don't have control over the market/customers. Nikon and Canon stumbled in the EUVL development race and the consequence was a massive loss in market share and complete loss of the EUVL market. In contrast look at Intel, they stumbled in the race to manufacture 7nm and below chips, but what are the consequences?

Intel still holds large market share and is no where near dropping out of the market or going bankrupt. Intel has a large 330 million person domestic market to fall back to if they stumble in their product development. TSMC's domestic market is 23 million, Samsung's domestic market is 51 million and Japan's domestic market is 126 million. Essentially the SK + TW + JP market is 130 million less than just the US market alone. Even worse is that SK, TW and JP companies are export competitors with each other.

Alarmists often daydream and talk about PLA missiles destroying TSMC fabs, but I think the real TSMC killer will be Intel and Samsung. Let's look at TSMC's future markets:
  • The US cut off TSMC's China market with sanctions and threats.
  • In the US, TSMC is going to face stiff competition against Intel (a preferred domestic US company).
  • In the rest of East Asia, TSMC is facing Samsung and Intel. SK will prefer Samsung and the US can apply pressure on Japan to work more with Intel.
  • Europe is actually the best prospect for TSMC since TSMC is one of ASML's largest customers and Europe doesn't have a major chip manufacturing competitor. But even then it's still competing against Samsung and Intel who are also ASML customers. This is also assuming that the EU does not eventually move/fund its own chip manufacturer.
Essentially TSMC has to always keep ahead of the competition or lose major market share with no domestic market to fall back on. This process won't happen overnight, but over the course of decades. Yes Chinese companies are behind on several technological aspects, but like Intel they have a large and secure domestic market to fall back to.
 
D

dfrtyhgj

Junior Member
Registered Member
ansy1968 said:
@tokenanalyst Bro another one from Lingbo Weibu, never heard this company, but having achieved a breakthrough what more of other unknown Chinese tech company.

Please, Log in or Register to view URLs content!

Please, Log in or Register to view URLs content!

Please, Log in or Register to view URLs content!


Please, Log in or Register to view URLs content!

Please, Log in or Register to view URLs content!

Please, Log in or Register to view URLs content!

353 views34 minutes a
Click to expand...
Warp speed development.
 
D

dfrtyhgj

Junior Member
Registered Member
Please, Log in or Register to view URLs content!

BOE's investment is for Yandong Microelectronics to build a 12-inch wafer fab using domestic IC manufacturing equipment and conducting production process verification of the equipment.

BOE hopes to improve domestic IC equipment's industrialization capabilities and market competitiveness, gaining greater control over the equipment supply chain. The investment will also enhance BOE's cooperation with chip design firms and wafer fab plants and stabilize its supply chain, said BOE's announcement.
 
FairAndUnbiased

FairAndUnbiased

Brigadier
Registered Member
Weaasel said:
I recently heard of nanoimprint lithography, which Canon has developed the equipment for, and which does not utilize projection lenses to make imprints of circuit patterns of a reticle unto photoresists resin layered unto semiconductor wafers, but instead "perfectly" positions the photoresist resin atop a wafer using inkjet technology and then also "perfectly" places and presses the reticle atop the photoresist and then upon the incidence of DUV light sources through the transparent portions of the reticle that impact the photoresists, circuit patterns capable of matching what can be achieved by ASML' s EUV machines (>10 nm) can be achieved.

Has China attempted any research and development into nanoimprint lithography? Does Canon have a good chance of commercially marketing its technology?
Click to expand...
Here's the problem: Optical lithography techniques like DUV and EUV allow for optical shrinking - you don't have to make a mask that has precision of 20-30 nm, you can make a mask that only has micron level precision then shrink it down with projection optics, something like a reverse microscope that takes a large image and shrinks it down. But for nanoimprint lithography it is a direct contact process - your master mask has to have the same dimensions as your final product, so to make something in the 20-30 nm scale, you have to already have 20-30 nm patterning capabilities... which you need optical lithography for (or electron beam lithography, which is still expensive).

dfrtyhgj said:
Please, Log in or Register to view URLs content!

BOE's investment is for Yandong Microelectronics to build a 12-inch wafer fab using domestic IC manufacturing equipment and conducting production process verification of the equipment.

BOE hopes to improve domestic IC equipment's industrialization capabilities and market competitiveness, gaining greater control over the equipment supply chain. The investment will also enhance BOE's cooperation with chip design firms and wafer fab plants and stabilize its supply chain, said BOE's announcement.
Click to expand...

Probably they're making display drive chips and power management ICs.
 
Status
Not open for further replies.
Top