Chinese semiconductor industry

Status
Not open for further replies.
tphuang

tphuang

Lieutenant General
Staff member
Super Moderator
VIP Professional
Registered Member
siegecrossbow said:
If it is not sustainable, Huawei would not have done it. They usually keep a very low profile to avoid being targeted. Just compare how easily they got ZTE versus how many hoops they had to jump through to ban Huawei.

All evidence points to the fact that Huawei and SMIC have a definitive workaround of the ban and have factored in the worst case scenario of US lawmakers going full retard and using the nuclear option. Otherwise they could easily have waited a year or two because contrary to popular belief, smartphone sale is not the core of Huawei’s business.
Click to expand...
Smartphone is a big deal. Huawei made $90b last year?

If they sell 60m phones next year, that will be $40b? Let's say mate 60 does really well and sell 17m units at $1000/phone, that's $17b alone of extra revenue.

At a time when Huawei has seen huge decline in it's profit while also paying out huge bonuses to keep everyone happy, you really don't think these additional sales matter?

Just looking at what I posted earlier, they are going to save $5b alone from not paying Qualcomm for soc and rf chips. All money it can spend on more r&d and build up cash position.

I think we are right to think that Huawei made this decision to go all in on Kirin due to yield and capacity data that we don't have access to. But 5g phones are also a big deal to them.
 
siegecrossbow

siegecrossbow

General
Staff member
Super Moderator
tphuang said:
Smartphone is a big deal. Huawei made $90b last year?

If they sell 60m phones next year, that will be $40b? Let's say mate 60 does really well and sell 17m units at $1000/phone, that's $17b alone of extra revenue.

At a time when Huawei has seen huge decline in it's profit while also paying out huge bonuses to keep everyone happy, you really don't think these additional sales matter?

Just looking at what I posted earlier, they are going to save $5b alone from not paying Qualcomm for soc and rf chips. All money it can spend on more r&d and build up cash position.

I think we are right to think that Huawei made this decision to go all in on Kirin due to yield and capacity data that we don't have access to. But 5g phones are also a big deal to them.
Click to expand...

My point is that they won’t do it just to make a petty statement if another round of sanctions can hobble them. The decision makers are not idiots. They know that the Americans will be irate, especially when elections are coming up. That has been factored into the decision making.
 
BoraTas

BoraTas

Captain
Registered Member
tphuang said:
Smartphone is a big deal. Huawei made $90b last year?

If they sell 60m phones next year, that will be $40b? Let's say mate 60 does really well and sell 17m units at $1000/phone, that's $17b alone of extra revenue.

At a time when Huawei has seen huge decline in it's profit while also paying out huge bonuses to keep everyone happy, you really don't think these additional sales matter?

Just looking at what I posted earlier, they are going to save $5b alone from not paying Qualcomm for soc and rf chips. All money it can spend on more r&d and build up cash position.

I think we are right to think that Huawei made this decision to go all in on Kirin due to yield and capacity data that we don't have access to. But 5g phones are also a big deal to them.
Click to expand...
SMIC banding with Huawei on something this impactful definitely has a reason. I am still in shock to be honest. Just the modem on its own is a big thing. The N+2 came out of nowhere and it is apparently ready for HVM. Considering that any SME that can do 14 nm is blanket banned to the country, I believe SMIC believes total de-Americanization to that point is very near. Because they are not going to risk their 40-14 nm business just to make a statement. Unlike Americans keep writing around, China is not into that.

Note: Most tools can not be rated for any specific node size, therefore the US blanket ban is broader than many would interpret.
 
R

Rafi

Junior Member
Registered Member
BoraTas said:
SMIC banding with Huawei on something this impactful definitely has a reason. I am still in shock to be honest. Just the modem on its own is a big thing. The N+2 came out of nowhere and it is apparently ready for HVM. Considering that any SME that can do 14 nm is blanket banned to the country, I believe SMIC believes total de-Americanization to that point is very near. Because they are not going to risk their 40-14 nm business just to make a statement. Unlike Americans keep writing around, China is not into that.

Note: Most tools can not be rated for any specific node size, therefore the US blanket ban is broader than many would interpret.
Click to expand...

Chinese officials seem very confident in the "tech" part of the Superpower rivalry, I get the impression that they have had or are very near to a "technical" breakthrough.
 
I

interestedseal

Junior Member
Registered Member
Alb said:
This could be very important as it might substitute components that are monopolized by MKS Instruments resolving another potential choke point. Any additional comments on this?
Click to expand...
There are at least eight companies in China working on RF generators for different plasma processes. They have applications in many processes like PECVD, PEALD, etch and ion implant, each with slightly different requirements.
 
olalavn

olalavn

Senior Member
Registered Member
tphuang said:
who said this piece?

If each wafer contains 450 good SoC for TSMC wafer of Snapdragon, then I may have to revise my total of total chips per wafer down (i did calculation based on 550 chips/wafer when it's possible this may only be 500 to 525, not sure TSMC yield here), which would further lead to lower # of good chip per SMIC wafer




Which again is the reason I am skeptical of the sub 50% yield number we have seen. Again, I am sure it was at that level for a long time, buy the sudden change in huawei, especially with the brazen dig at raimondo tells me that they have seen an improvement recently in kirin production. Otherwise, how do you tell supply chain to prepare for 15 to 20m mate 60 phones and then tell stores to not worry about supply? And that's just tip of the iceberg. So many new phones out that will need Kirin chips.

And now they get a full ban, they must have done some calculations and think that smic + whatever they stocked up is sufficient to last until whenever smic can ramp up production. My guess sometimes over next year.

My calculations was that even 20k wpm of 7nm isn't enough for Huawei itself let alone all the other OEMs in china.

And that number needs to continue to increase, because they will have to keep going advanced packaging and larger combined die size to deal with the gap in process node.

So one would think that either Huawei just went crazy here or they are seeing a way through. Which means more finfet capacity+ higher yield expected going forward.
Click to expand...
Please, Log in or Register to view URLs content!

he write this... they are using a lot of sub-7nm processes to make better chips...
next year, the 60m/year milestone may be possible..
 
T

tokenanalyst

Brigadier
Registered Member
BoraTas said:
SMIC banding with Huawei on something this impactful definitely has a reason. I am still in shock to be honest. Just the modem on its own is a big thing. The N+2 came out of nowhere and it is apparently ready for HVM. Considering that any SME that can do 14 nm is blanket banned to the country, I believe SMIC believes total de-Americanization to that point is very near. Because they are not going to risk their 40-14 nm business just to make a statement. Unlike Americans keep writing around, China is not into that.

Note: Most tools can not be rated for any specific node size, therefore the US blanket ban is broader than many would interpret.
Click to expand...
I do agree that something is going on under.

People tend to underestimate that a lot of "mature" tools including many Chinese made tools are used across multiple process nodes from mature to advance and multiple layers. For sake, I-Line with mercury vapor lamps scanners are still used in advance chip manufacturing maybe not for the critical features but without them would be really expensive to etch the metal layers of a chip.

tokenanalyst said:
For example in this paper:
View attachment 112391
ICRD used ASML immersion scanner, a Naura etching tool and Piotech ALD tool in the design of this process. So this demonstrate deposition and etching tools used in mature process can be used in more advanced process. But I guess it could come with some yield loss.

View attachment 112392View attachment 112394

Please, Log in or Register to view URLs content!
Click to expand...
People may say "NO, is just a lab experiment" "muh HVM", in reality lot of China semiconductor processes are first developed in collaboration with the ICRD and a lot of tools are certify by them, even ASML has a collaboration with the ICRD. So there is that, this idea that Chinese tools are just for very mature process need to be revised, more and more Chinese tools are going to get verified for advance process nodes.

Please, Log in or Register to view URLs content!

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

tokenanalyst

Brigadier
Registered Member
In the research literature there is evidence that SMIC has been trying to overcome the shortcomings of domestic tools for their use in advanced processes.

tokenanalyst said:

Extending the Application of Capacitively Coupled Plasma Etching Tools to the Front-End-of-Line Fin-Cut Etching Process for FinFET Mass Production.​

Semiconductor Manufacturing International Corporation

Abstract​

As FinFET scales down to advanced technology nodes, Self-aligned Quadruple Patterning (SaQP) scheme is widely deployed for Fin formation, where two mandrels (MD1 and MD2) and corresponding mandrel spacers (MD1SP and MD2SP) are patterned for accurate CD transferring. The Fin-cut etching process of the MD2SP Cutting is very critical for the patterning of the Standard Cell and the SRAM patterns. This work studies the application of the Front End of Line (FEoL) MD2SP Cutting process in both inductively coupled plasma (ICP) and capacitively coupled plasma (CCP) etchers. The mismatch between the two tools is observed and discussed in detail. For CCP tools, a degradation of line width roughness (LWR) and increase of critical dimension (CD) loading between dense and isolated patterns are observed. Compared with the ICP etchers, the CCP etchers have lower plasma dissociation and stronger physical bombardment of ions. Furthermore, it is harder to control the plasma density and directionality separately. To increase CCP tool applications in FEoL processes, we introduce a Silicon coating process to modify the photoresist (PR) during the CCP etching to harden the surface of the PR and enhance the LWR performance. Furthermore, the application of atomic layer etching (ALE) technology ensures the CD loading between the dense and isolated patterns. With the application of our study, we can broaden the application of CCP tools in FEoL for FinFET mass production.


Please, Log in or Register to view URLs content!

Looks like SMIC has manage to solved some issues on the etched resolution of patterns made with capacitive couple plasma etchers.
AMEC and Naura etchers?
Click to expand...
 
Status
Not open for further replies.
Top