Good thing I had put what I was drinking down before I read this, I would hate to short circuit my keyboard. They're comparing SMIC's volume production to their fantasies.
Germany Carl Zeiss, heart of Dutch ASML Lithography Equipment.
- Thread starter tidalwave
- Start date
- Status
A very good Phd thesis on MOPA(Master Oscillator Power Amplifier) Laser architecture which can be used in EUV light generation.
From the PHD Thesis demonstrates the MOPA achitecture
The system comprised a pulsed seed laser at 1064nm and two stages of amplifiers called pre-amplifier and a power amplifier, both based on Nd:YVO 4 gain medium in the bounce geometry configurations.
1)The pulsed seed source used in this system was a semiconductor laser diode. The laser diode was connected to a pulsed electrical driving circuit for gain-switched operation. The electrical drive circuit was capable of generating an optical pulse durations from 3.5ns to over 10us with adjustable pulse repetition rate, ranging from single shot up to 2MHz.
*)The wavelength of the laser diode was therefore tuned using a diffraction grating.
2)The pre-amplifier and power amplifier were both based on 1.1 at. % Nd:YVO 4 crystal using bounce geometry configuration. The crystal was side-pumped by a laser diode bar emitting at 808nm, which was anti-reflection (AR) coated for the pump wavelength. The two end laser faces were AR coated at 1064nm. The heat was removed from the top and bottom faces of the crystal using water-cooling
system,
From the PHD Thesis demonstrates the MOPA achitecture
The system comprised a pulsed seed laser at 1064nm and two stages of amplifiers called pre-amplifier and a power amplifier, both based on Nd:YVO 4 gain medium in the bounce geometry configurations.
1)The pulsed seed source used in this system was a semiconductor laser diode. The laser diode was connected to a pulsed electrical driving circuit for gain-switched operation. The electrical drive circuit was capable of generating an optical pulse durations from 3.5ns to over 10us with adjustable pulse repetition rate, ranging from single shot up to 2MHz.
*)The wavelength of the laser diode was therefore tuned using a diffraction grating.
2)The pre-amplifier and power amplifier were both based on 1.1 at. % Nd:YVO 4 crystal using bounce geometry configuration. The crystal was side-pumped by a laser diode bar emitting at 808nm, which was anti-reflection (AR) coated for the pump wavelength. The two end laser faces were AR coated at 1064nm. The heat was removed from the top and bottom faces of the crystal using water-cooling
system,
SMIC still a small player with yearly revenue of only $3bilion, it cannot handle all Huawei's volume.
Huawei has to prepare one day TSMC may not supplying manufacturing due to US pressure.
It has to get involved into chip manufacturing like samsung. And be a full blown semiconductor company too, selling all its designed and manufactured chips.
Huawei is a big player with $100billion yearly revenue. It can form a fab and license from SMIC and have SMIC engineers step walk its own engineers in chip manufacturing.
Maybe Huawei can also morph into big fundamental science player like a Applied Material, KLA Tencor, ASML in develope semiconductor material/equipments. That would be the ideal situation,.
That way it can fully control its own destiny.
Huawei has to prepare one day TSMC may not supplying manufacturing due to US pressure.
It has to get involved into chip manufacturing like samsung. And be a full blown semiconductor company too, selling all its designed and manufactured chips.
Huawei is a big player with $100billion yearly revenue. It can form a fab and license from SMIC and have SMIC engineers step walk its own engineers in chip manufacturing.
Maybe Huawei can also morph into big fundamental science player like a Applied Material, KLA Tencor, ASML in develope semiconductor material/equipments. That would be the ideal situation,.
That way it can fully control its own destiny.
Maybe taking on a Samsung path is ideal. I don't think it'd be wise to be a much more fundamental Science player like ASML and Applied Material. Let that role be done by others. Huawei should not be burdened with too many a task. I'd like it to have its own fabs. But like TSMC...having fabs and making own lithographic equipments are different things. Huge capital investment needs to be ploughed into Foundational Science players by the government. Huawei can also pitch in and buy their products.
Currently the only Chinese company makes lithography equipment is SMEE but no one buys its product because no one trust them,
If Huawei makes the equipments and definitely be used in large scale for own production.Huawei has the scale to make the breakthrough.
I rather see a impactful fundamental science company emerges in China than anything.
It represent the top food chain and rest in the hands the Germans, Dutch, Americans, and Japanese.
In fact Japan recently sold panasonic semiconductor to taiwan. Japan doesn't want to make chips anymore instead focus on materials and equipments.
Taiwan and Korea make alot of chips but they are only application users of those tools.
China cannot depend on western tech in any ways. I rather see Huawei license out its OS like Google and sell its own manufactured chips and focus more on fundamental science. Huawei is the best candidate to make that transition because it has scale and ability.
I agree that Huawei has to prepare for the day TSMC stops supplying, but I would disagree on Huawei having to get into semiconductor manufacturing like Samsung.
From a strategic viewpoint, a world-class Chinese semiconductor company would need:
1. As much unrestricted access to overseas semiconductor technology as possible.
2. As large a customer base as possible, in China and globally. This funds development and justifies the scale of investments.
Korea's economy is dominated by a few chaebols, which means there isn't space for a independent pure-play semiconductor foundry like TSMC/SMIC to emerge.
In comparison, the Chinese economy can and does support numerous small/medium/large companies which can specialise in certain areas and be more efficient and faster in product development.
Unlike Samsung, Huawei is already on the Entity List which will impede its acquisition of the latest semiconductor technology.
And it's only going to get worse, which means the potential Huawei semiconductor customer base will become more limited.
Remember Samsung can only justify its semiconductor investments by selling components to a global customer base.
Plus Huawei semiconductor's potential customers would also be deterred, because it would be a competitor for many products.
SMIC is already 1 of only 5? companies that has 14nm FINFET process technology, and should be able to ramp this up capacity to whatever level Huawei requires in the next few years. Plus SMIC can sell to any customer in China or overseas.
Also note Intel is currently stuck at 14nm.
And that Huawei is a minority shareholder in the SMIC subsidiary developing 14nm.
And given the shortage of semiconductor engineering talent, spreading SMIC personnel even further to Huawei doesn't make sense.
I also reckon demand can only support 1 Chinese semiconductor company with process nodes smaller than 14nm.
So it wouldn't make sense for Huawei to get into this at the expense of SMIC.
The same sort of logic applies to Huawei getting into semiconductor materials research and toolmaking.
Huawei being a customer and supporting other companies in semiconductor development makes more sense.
One other interesting point.
It takes 4months to process a 14nm silicon chip. So it's inherently slow going when testing different recipes.
As much as I support Huawei in its fight against unjustified attack from USA, we should also be aware of two things.
Huawei's leader Ren Zhengfei is more clear-headed by saying "we are JUST a company, please don't put us (too much) to the spot as national hero". I think, some times layman's overly desire and expectation on Huawei would hurt it rather than help it. The bottom line is one can not build a single battleship over 1,000,000 tone to defeat all navies on the globe, it has to be a fleet of 1000 much smaller ships.
- Huawei is a collective ownership company, SMIC is unrelated to Huawei. There is no justification of SMIC to be "sacrificed" for Huawei.
- Huawei is not the only Chinese companies needing high end semiconductors. There is no reason for Chinese state to favorite Huawei over others including Huawei's competitor ZTE which the state has a big share.
Huawei's leader Ren Zhengfei is more clear-headed by saying "we are JUST a company, please don't put us (too much) to the spot as national hero". I think, some times layman's overly desire and expectation on Huawei would hurt it rather than help it. The bottom line is one can not build a single battleship over 1,000,000 tone to defeat all navies on the globe, it has to be a fleet of 1000 much smaller ships.
Right now it's reaching point of state of emergency. Alot of Chinese tech companies under US sanction not just huawei. If Trump reelected it's gonna get worst. SMIC would also be sanctioned.
Only the end result counts, doesnt matter how to do it. And most importantly everyone has to pitch in. Huawei is a big player and definitely can pitch in directly or indirectly.
What China needs to do now is going to second market and purchase as many equipments and materials as possible. IF the big axe comes down. hopefully those equipments are enough to last until self developed ones come on board.
Right now conventional thinking is not enough to fight the firestorm
Every coin has two sides. On the one side, ASML and Netherlands are under pressure (leverage) of Trump, so media especially Trump's media want to paint the picture in a way as if ASML will not deliver. But there is the flip side of the same coin. Netherlands have trade surplus (if I am right) in trade with China, a leverage of China. ASML's machine is not big volume commodity, they sell less than 5 (in that range) per year? Or less. Loosing SMIC's contract for many years is a big blow to ASML itself. And all this for what? Netherlands has its own interest different from USA, is Trump going to compensate Netherlands' loss in China? Impossible, just look at the damage of Thaad to SK, Trump added salt to that wound by further squeezing SK on trade deal. Look at Japan, what Japan get in return from fighting China on Obama's behalf? Punched in the face by Trump, now we are seeing Japan begin to turn around.
A long delay (by a year) of ASML delivery may pose a serious damage to SMIC's position since the life cycle of semiconductor is short, but such delay also carries a significant and long term negative impact on bilateral relationship between the Netherlands (and EU as a whole) and China on strategical level, equivalent to a new iron curtain (tech and trade). Is EU willing and ready to go that path? I doubt.
A long delay (by a year) of ASML delivery may pose a serious damage to SMIC's position since the life cycle of semiconductor is short, but such delay also carries a significant and long term negative impact on bilateral relationship between the Netherlands (and EU as a whole) and China on strategical level, equivalent to a new iron curtain (tech and trade). Is EU willing and ready to go that path? I doubt.
SMIC is a small player and not significant buyer of ASML EUV machine, it only purchased one, cost $120million, the entire year's profit gone.
look at TSMC purchased around 16 EUV . Samsung is also a big purchaser.
- Status