Chinese semiconductor industry

ZeEa5KPul said:

I agree with you for the most part on the process of de-Americanization of supply chains. There are huge incentives for countries like Japan and Korea to develop "Certified 100% America Free" technology products for the Chinese market - however, that assumes the US is willing to play by at least some rules and not sanction these countries for supplying advanced technology even if it contains no US components or IP whatsoever. Given the perfidy and thuggery we've seen from the US, there's no reason to assume any such restraint on its part. Hell, Japan faced sanctions in the '80s for developing such technology itself! Any objection that that would be against the law (whose law?) is futile. The US is dying and it's determined to play Samson and pull the temple down on itself and everyone else, that's just a fact. Therefore, there must be total Chinese self-reliance (not just that, but world leadership) in core technologies like semiconductors no matter what it takes.

Click to expand...

ZeEa5KPul said:

I also want to question you a little on your timelines for DUV and EUV introduction. You say it'll take time for the first commercial Chinese DUV machine to reach 14nm (even 28nm), let alone 7nm. Ordinarily I would agree, given the lackadaisical pace of Chinese research in lithography prior to US sanctions, but now I simply cannot see anything but a full-on war emergency effort to get this technology to the cutting edge. It would be a mistake for Chinese DUV to spend a day at 28nm - straight to 14nm and in months transition to 7nm. As for EUV, I'm not exactly sure what the Changchun Institute announcement in 2017 is about, but if it's a complete tech demonstrator then I expect the first generation of commercial machines to appear, J-20 like, in the next year or two (or three, just to cover my rear).

Click to expand...

superdog said:

If they don't contain US technology then there is no part to sanction. If you mean through other channels such as canceling unrelated agreements or exerting political pressure, yes of course we can always expect the US to do that. They have already been doing that in persuading other countries to not buy Huawei. Doesn't mean it will always work, and every time the US tried it is draining its political capital. I get your concern that this could make it unreliable if part of the supply chain is completely dependent on a 3rd country, and it is important to get hold of core technologies at home, I agree.


We know that the domestic DUV (immersed) machine is not yet ready. Rumor says it will begin production testing next year, let's say it will be delivered early by 2nd half of 2020, setting it up, passing test production, working out all the bugs, making it commercially viable for 28nm production, and then push towards 14nm, as well as making enough machines to fulfill capacity......even if they didn't encounter any big obstacle, these will still take time. I can't (and I didn't) say exactly how long, but if all these only took 2 years I'd be more than happy.

For these kind of projects you have to take the scientific approach, trial and error takes time, knowledge accumulation takes time, training people takes time, throwing money at it can't accelerate this process proportionally. Skipping steps could also make you lose more time instead of gaining it.

Click to expand...

superdog said:

We know that the domestic DUV (immersed) machine is not yet ready. Rumor says it will begin production testing next year, let's say it will be delivered early by 2nd half of 2020, setting it up, passing test production, working out all the bugs, making it commercially viable for 28nm production, and then push towards 14nm, as well as making enough machines to fulfill capacity......even if they didn't encounter any big obstacle, these will still take time. I can't (and I didn't) say exactly how long, but if all these only took 2 years I'd be more than happy.

Click to expand...

superdog said:

For these kind of projects you have to take the scientific approach, trial and error takes time, knowledge accumulation takes time, training people takes time, throwing money at it can't accelerate this process proportionally. Skipping steps could also make you lose more time instead of gaining it.

Click to expand...

ZeEa5KPul said:

I see. To put the question bluntly, how much danger do you think is China in? Can its chip stockpile and already imported DUV machines hold the line until SMEE's SSB800 is ready? If not, how big of a gap do you think there'll be between the end of the stockpile and SSB800 producing?

Click to expand...

hullopilllw said:

Whoever bought over HiSilison will be added into the entity list.

You need to understand the underlying intent of US State Dept here, it is not really only against Huawei, but China's ascend in the semiconductor sector as a whole.

Click to expand...

superdog said:

For China? I think the entity list is not much of a danger but more like a stimulation. It is a big incentive to speed up the semiconductor industry upgrade, doing a sanction now is way too late to take away China's ability to upgrade. National security is not affected, as the defense sector has been sanctioned long ago. Also note that this is just a unilateral US sanction on some selected companies, not something like the UN sanction on Iran (which would be impossible to pass on China), the scope of influence is much more limited. It's not like China suddenly wouldn't be able to make or buy chips.

Top concern for China right now is keeping the virus under control, revive the economy, and then think about how they should act as a major economy that revived early while the US is still in deep trouble. Lots of things could be done during this period. China have a lot of cards to play so I'm not worried about it.

For affected companies like Huawei this is of course a bigger challenge, if at the end no compromise is reached and everything strictly sanctioned, they may have to hibernate their consumer electronic business and make adjustment to other departments, this will have a big impact to their business plan. But even in this situation I'm confident that A) Huawei's survival is not an issue, B) the crippling effect will be temporary, and C) the impact to China's 5G development will be fairly limited.

Click to expand...

AndrewS said:

I agree.

The economic damage of the Virus First Wave already looks far worse in the USA than in China.
That should mean China has a stronger hand in the coming months.

China had a 6.8% decrease in China GDP in the 1st Quarter, but the indicators are that China is back up and running at almost 100% now after the lockdown, even with the few localised outbreaks we've seen.

Deutsche Bank is forecasting a 40% decline in US GDP in the 2nd Quarter.
Even taking into account many attenuating factors, the economic impact in the USA is far greater.

And going forward, it does look like China will be better able to track and control future outbreaks.
That is partly because the initial Chinese lockdown was much more comprehensive so that the virus was virtually eliminated.
And they've developed the monitoring, tracking and testing systems to prevent large outbreaks.

In comparison, there are still 25000 confirmed cases per day in the USA, yet a number of states are already reopening.
It's inevitable that there are going to be a lot more outbreaks for the rest of the year in the USA. Source below.

Please, Log in or Register to view URLs content!

Click to expand...

AndrewS said:

It's possible HiSilicon could be added to the Entity List, but they would have to come up with yet another justification.
In which case, the situation is no different than if HiSiliicon is part of Huawei.

And I said "spun off as an independent entity" or "sold to Unisoc/Tsinghua"

If HiSilicon can supply to everyone, there is still a net overall benefit to the Chinese semiconductor industry,

Click to expand...