Nothing changes the fact that it will take some time to design alternatives to US technology.
But given enough time, Chinese companies will be able to come up with alternatives in every area.
Germany Carl Zeiss, heart of Dutch ASML Lithography Equipment.
- Thread starter tidalwave
- Start date
- Status
Nothing changes the fact that it will take some time to design alternatives to US technology.
But given enough time, Chinese companies will be able to come up with alternatives in every area.
So from what I can see, Huawei is/was sourcing chips as follows:
TSMC: 14nm (15% US content)
TSMC: 7nm (9% US content)
TSMC: 5nm (? US content)
SMIC: 14nm (? US content)
Samsung: 6nm (? US content)
Any ideas on the content levels of SMIC or Samsung?
Presumably Samsung will also be reducing their US content levels
It looks like Samsung semiconductor ships more chips to China than the rest of the world, and they definitely do not want to be caught out by the Entity List.
And we know SMIC will be reducing their US content as much as is feasible.
Plus it's public knowledge that TSMC are resisting the effects of any Entity List restrictions.
Then it's also useful to know that TSMC owns 8% of SMIC, and has the option to buy 10% more.
And that Mediatek 7nm 5G chips made by TSMC should also be available to Huawei.
So in summary:
1. Huawei looks like it will have a diversified selection of chip suppliers for the coming year
2. Time is the most important thing for Huawei and the overall semiconductor industry. That time is needed in order for everyone to remove US technology content.
TSMC: 14nm (15% US content)
TSMC: 7nm (9% US content)
TSMC: 5nm (? US content)
SMIC: 14nm (? US content)
Samsung: 6nm (? US content)
Any ideas on the content levels of SMIC or Samsung?
Presumably Samsung will also be reducing their US content levels
It looks like Samsung semiconductor ships more chips to China than the rest of the world, and they definitely do not want to be caught out by the Entity List.
And we know SMIC will be reducing their US content as much as is feasible.
Plus it's public knowledge that TSMC are resisting the effects of any Entity List restrictions.
Then it's also useful to know that TSMC owns 8% of SMIC, and has the option to buy 10% more.
And that Mediatek 7nm 5G chips made by TSMC should also be available to Huawei.
So in summary:
1. Huawei looks like it will have a diversified selection of chip suppliers for the coming year
2. Time is the most important thing for Huawei and the overall semiconductor industry. That time is needed in order for everyone to remove US technology content.
You should thank Tidalwave, he is trying to help Xi.
coolieno99
Junior Member
The VC group asked me to give a cost estimate to buy equipments and parts and setup a demo in the lab.
Well, it's not simple. I will break it down into 3 domains. 1 is the Laser unit that generate EUV light. 2 is the mirror needed for the optical system. 3 is the overall optical system design on how the EUV light propograte.
I suggest they would start on 2, acquire the manufacture knowhow of the mirror. 3, i already got that software and I will give to them along with application files that use this software. I suggest to them to partner up with a univeristy and let the whole group of engineering student play with software until they very proficient in it and start simulate the optical system. The software change change to EUV light based on wavelength and mirror stackup for simulation.
For 3, it's bit hard, meanwhile I acquire the parts to do a proof of concept, master oscillator and amplifier output. Not exactly the same output as the real EUV light generation and but demo of the architecture needed.
For 2, acquiring the knowhow for EUV mirror manufacturing. Since its silicon substrate and on top of it, its Molybenum and Silicon each one is about 3.4nm thick. There 40 layers of them.
I surveyed the chinese market there are DC magnetron Sputtering system for deposition. But I not sure how reliable they are. I may suggest them just use a Fab service like SMIC, HuaHong for making mirror, its like making the chip, dimension wise.
Next is milling for roughness of surface of the mirror down to the spec. Milling would use focused Ion Beam (FIB) tool, each system in US cost $500K. There are couple vendors in Beijing capable of doing Electron Scanning Electron Microscope(SEM), but its not quite FIB. FIB uses Ga+ ion instead of electron to bombard the surface of the mirror to mill or chip the material an atom at a time.
I would suggest to pay the engineering service and cost of one unit to beijing vendor to modifiy its SEM into a FIB. Once they do that, the VC group would retain the IP to the FIB since they pay for engineering service.
Next is the Interferometer + Microscope to profile the mirror surface, the laser interferometer would measure the peak and valley of surface of the mirror. Again, I would suggest to pay for engineering service to acquire such system., There are already chinese vendors capable of doing interferometer , microscope but not combining together. And also write the software for this tool.
So, i suggest them to pay for the engineering service for these two tools and therefore they acquire two core technologies for EUV mirror.
so, this is just starting point of ambitious project that would change the supply chain landscape in China. I don't suggets and go the route of buying ready made western tools try to do something because you don't change your underlying supply chain, you will never cure the fundamental problem.
Well, it's not simple. I will break it down into 3 domains. 1 is the Laser unit that generate EUV light. 2 is the mirror needed for the optical system. 3 is the overall optical system design on how the EUV light propograte.
I suggest they would start on 2, acquire the manufacture knowhow of the mirror. 3, i already got that software and I will give to them along with application files that use this software. I suggest to them to partner up with a univeristy and let the whole group of engineering student play with software until they very proficient in it and start simulate the optical system. The software change change to EUV light based on wavelength and mirror stackup for simulation.
For 3, it's bit hard, meanwhile I acquire the parts to do a proof of concept, master oscillator and amplifier output. Not exactly the same output as the real EUV light generation and but demo of the architecture needed.
For 2, acquiring the knowhow for EUV mirror manufacturing. Since its silicon substrate and on top of it, its Molybenum and Silicon each one is about 3.4nm thick. There 40 layers of them.
I surveyed the chinese market there are DC magnetron Sputtering system for deposition. But I not sure how reliable they are. I may suggest them just use a Fab service like SMIC, HuaHong for making mirror, its like making the chip, dimension wise.
Next is milling for roughness of surface of the mirror down to the spec. Milling would use focused Ion Beam (FIB) tool, each system in US cost $500K. There are couple vendors in Beijing capable of doing Electron Scanning Electron Microscope(SEM), but its not quite FIB. FIB uses Ga+ ion instead of electron to bombard the surface of the mirror to mill or chip the material an atom at a time.
I would suggest to pay the engineering service and cost of one unit to beijing vendor to modifiy its SEM into a FIB. Once they do that, the VC group would retain the IP to the FIB since they pay for engineering service.
Next is the Interferometer + Microscope to profile the mirror surface, the laser interferometer would measure the peak and valley of surface of the mirror. Again, I would suggest to pay for engineering service to acquire such system., There are already chinese vendors capable of doing interferometer , microscope but not combining together. And also write the software for this tool.
So, i suggest them to pay for the engineering service for these two tools and therefore they acquire two core technologies for EUV mirror.
so, this is just starting point of ambitious project that would change the supply chain landscape in China. I don't suggets and go the route of buying ready made western tools try to do something because you don't change your underlying supply chain, you will never cure the fundamental problem.
this is synchrotron,
it cannot be used to make chips because look at the size of the structure, not viable for commercial production. Its for scientific study.
antiterror13
Brigadier
wowww, you are a top consultant. I've never seen a top consultant before boasting the detail of the discussion in a forum ... you are a top man ... keep a good work
It seems you are super efficient top consultant as well, still have a lot of time spending time in this forum ... well done
The information you are suggesting here, are they not company internal trade maters (trade secrets)? Would they be happy to share their company informations with their competitors inside or outside China?
antiterror13
Brigadier
- Status