Array
(
    [content] => 
    [params] => Array
        (
            [0] => /forum/threads/china-blames-us-for-new-blow-to-strategic-high-tech-industry.21001/page-2
        )

    [addOns] => Array
        (
            [DL6/MLTP] => 13
            [Hampel/TimeZoneDebug] => 1000070
            [SV/ChangePostDate] => 2010200
            [SemiWiki/Newsletter] => 1000010
            [SemiWiki/WPMenu] => 1000010
            [SemiWiki/XPressExtend] => 1000010
            [ThemeHouse/XLink] => 1000970
            [ThemeHouse/XPress] => 1010570
            [XF] => 2021770
            [XFI] => 1050270
        )

    [wordpress] => /var/www/html
)

China Blames US for New Blow to Strategic High-Tech Industry

I wonder if they see a silver lining though. China has seem dead set on being fully technologically self-reliant. The lack of sanctions wouldn't have had a huge impact on their efforts to copy DUV and EUV technologies. In some ways it would slow them down (less urgency), in some ways it would speed up the efforts (getting more experience with how EUV machines work, perhaps some aid to reverse engineering efforts).
Had the US waited to apply the sanctions until it had an actual conflict with China it would have been much more effective. In this way China will have its own counter-measures in place by the time a conflict does happen.
It is also one thing to ban supply of tools you had not supplied before like EUV, and quite another to cut supply and parts of equipment you had sold already breaking established contracts. There is simply no excuse for this. Not when the US isn't in a conflict with China.

In the long term this isn't good for ASML. As soon as China has a competing technology they will massively subsidize their industry to drive out competitors, as we've seen them do in other areas. That is if they still have the economic power to do so a decade or two down the line.
Of course the Chinese will. China has the world's largest semiconductor market. It has enough critical mass to be completely indigenous if it chooses to do so.

I'm not so sure actually. The Chinese model seems to be near a breaking point. If this was any other country their chip efforts would be considered insane. No chip company in a free market could survive making 7nm chips without a strong international market with healthy profit margins.
The Chinese market is just that big. Maybe 20-30% of world semiconductor demand internally, bigger than the US, and more like 50% if you count for semiconductors embedded in Chinese exports.

How many chip industry bankruptcies can the country handle? This is all funded by a staggering amount of debt in a debt-model that we know with full certainty is completely unsustainable
The Chinese are competing against incumbents with depreciated fab assets. This is just par for the course if they want to have their own semi industry and not be vulnerable to chip sanctions in the future.

Remember that SMIC was built in part by poaching TSMC talent.
The founder was from a Taiwanese company which was a competitor with TSMC. Several TSMC engineers (which developed the 28nm process) moved to Samsung (where they developed a 14nm FinFET process) and later to SMIC.

It's hard enough to recruit talent to China already (no option for a non-Chinese to build a life there long term). If they can't pay top dollar for talent it'll be even harder.
They have a lot of Chinese expats all over the world which they can call back to China if they need talent. They also seem to be successful in attracting other East Asian scientists including from Japan.

If they can't even build up a sustainable domestic economy around something as simple as a ballpoint pen tips, how will they build a completely self-reliant chip manufacturing economy? I think they will try, but I think there's a significant chance they'll break their own economy trying.
You learned the wrong lesson I suppose. The Chinese did make their own ballpoint pen tips in the end. And ballpoints are not a particularly simple thing to make because of the tolerances involved.
 
Last edited:
Perhaps a really dumb question, but for military applications is a “3nm” chip really giving you that much benefit over a “7nm” chip these days? (or perhaps a 2nm vs 5nm which is what it looks like it’ll be in 2 years).

I can’t see weapons systems benefitting that much (is 3nm suddenly going to run AGI when 7nm can’t?), and space hardware needs mature chips for reasons. For large scale computing I can certainly see benefits of 3nm but China has no problem just simply solving the issue with brute force (just build 4x as much or whatever, and power it with whatever fuel source can be made available).
Perhaps, autonomous drones and robots for warfare would require large area and low power SoCs (high density libraries/nodes) that only leading edge nodes could provide.
Idk, if they could prevent Intel from spinning off IFS like they didn't allow the Tower Semiconductor merger.
The PRC are already have mastered fighter jet turbofan engines (it took them decades to do that). With tons of skilled engineers, investment, industrial espionage, and anything else they can muster, I'm sure the time to completely master DUV and eventually EUV isn't that far off: it'll take far less time than people think.
 
Perhaps, autonomous drones and robots for warfare would require large area and low power SoCs (high density libraries/nodes) that only leading edge nodes could provide.
Idk, if they could prevent Intel from spinning off IFS like they didn't allow the Tower Semiconductor merger.
The PRC are already have mastered fighter jet turbofan engines (it took them decades to do that). With tons of skilled engineers, investment, industrial espionage, and anything else they can muster, I'm sure the time to completely master DUV and eventually EUV isn't that far off: it'll take far less time than people think.

if china still has this kind of '711' scientist occuping the top R&D positions, people can just relax. it's just wasting it's money on useless effort.
1726705860934.jpeg
 

Attachments

  • Screenshot 2024-09-18 135845.jpg
    Screenshot 2024-09-18 135845.jpg
    91.6 KB · Views: 104
Last edited:
I'm not sure what [Huawei] "mastering ARM design" means here and that it's really anything that much to brag about. It usually means slight customisation/configuration and implementation from RTL. Compared to many other leading edge chip design tasks, not of the highest difficulty. Nor one where the skills are that difficult or time consuming to master - or where experienced people can't be found and hired.

To the best of my knowledge Huawei never had an ARM architecture licence (like Apple for example). Happy to be corrected here. That's a whole different skill level.
Simply not true. Huawei's HiSilicon develops its own ARM core designs just like Apple. The Taishan core family for example is used as the big cores in the Kirin 9000S. They also develop their own GPUs named Maleoon. Also used in the Kirin 9000S. The big Taishan CPU cores are also used in their Kunpeng server processors.

In addition to CPU and GPU cores HiSilicon develops AI NPUs and their own 5G modems. The only thing they have not designed thus far is the small smartphone cores. They originally developed the large cores for their server CPUs and then used them in smartphones later.
 
My two cents: ASML-CEOs are not happy, but the US still determines the shape and intensity of the chip-war (so far), it may be a long war......It will be crucial how BRICS+ will respond to the patent-issues, if China is not allowed to buy this EUV and DUV technology, at some point I think they will master the technology, breaking IPR or not, if needed "hiding" behind the same arguments of national security......
The natural consequence of this move is EU countries now asking why hundred+ US companies are allowed to export to China, when US pressures a foreign one on export control. It's not only litho, but every other piece of high-tech equipment.

How NL sees it: US chose the least expensive method for itself to deprive China of microchips
 
Simply not true. Huawei's HiSilicon develops its own ARM core designs just like Apple. The Taishan core family for example is used as the big cores in the Kirin 9000S. They also develop their own GPUs named Maleoon. Also used in the Kirin 9000S. The big Taishan CPU cores are also used in their Kunpeng server processors.

In addition to CPU and GPU cores HiSilicon develops AI NPUs and their own 5G modems. The only thing they have not designed thus far is the small smartphone cores. They originally developed the large cores for their server CPUs and then used them in smartphones later.
So let's get this straight.

You are saying that Huawei has an ARM architectural licence. Correct ?

And that is still the case today ?
 
So let's get this straight.
You are saying that Huawei has an ARM architectural licence. Correct ?
And that is still the case today ?
ARM is not revealing the name of its new architectural licensee. The company has eight publicly announced 64-bit architectural licensees: Applied Micro, Broadcom, Cavium, Apple, Huawei, Nvidia, AMD and Samsung. It also has another seven publicly announced 32-bit architectural licensees, of which five – Marvell, Microsoft, Qualcomm, Intel and Faraday – do not have a 64-bit licence.
 
Back
Top