China cooks covert chips, recruits global geeks to dodge US restrictions China is looking to hire engineers from other nations to bolster local chip production in the face of crippling US sanctions, with Huawei – one of the country's largest tech giants – said to be creating a network of covert semiconductor plants across the country. As Washington tries to curtail China's access to advanced …
Will an unexpected consequence of Western policy be that China will develop chip manufacturing expertise that is on a par with that achieved by machines from ASML and the ability to design them on a par with Intel, AMD, ARM, ... ?
At that point China will point two fingers at us and ship chips to BRICS & other nations at a price that undercuts what the West can make them at.
Probably it is a certain inevitability, alain williams, and much sooner than was never before imagined possible too.
It is a truly unfortunate and unedifying, self-destructive trait that there is an unpleasant abiding systemic arrogance in Western leaderships that presuppose the East incapable of superior leading advancement without their virtual aid and practical assistance.
However, notwithstanding that, methinks a much greater fear for the West to tie themselves up in knots over and run around like headless chickens to fail to prevent is the secure covert and clandestine Eastern application of stealthy operational software programs/insider projects in chips vital to Western national, international and internetional security ...... and which they might realise are active but which they can do absolutely nothing about because of the untold damage removal of the access vehicle would present.
Before there was cheap Chinese crap, there was cheap Japanese crap.
They started out slavishly copying western products. But with the corners cut and prices cut even more.
Then they worked out that improving the quality would save them money: returns are expensive to deal with.
Finally, they started slavishly copying designs the west hadn't made!
Will an unexpected consequence of Western policy be that China will develop chip manufacturing expertise that is on a par with that achieved by machines from ASML
That would take at least a decade, minimum. EUV is the result of 20 years of research, and there are hundreds of components in it that had to be invented from scratch to make it work. That's why even if China stole the blueprints for ASML's EUV scanner and had a working model to compare to, they would still require at least a decade to replicate it.
There are known ways to achieve what EUV is capable of that China has access to, like e-beam. The problem is that throughput with e-beam is a tiny tiny fraction of what is required for mass production, but for China's military needs that may not matter. But absent a massive breakthrough it wouldn't serve the commercial needs of a company like Huawei.
I will wager that long before twenty years are up EUV will not be necessary.
I wonder how far away we are from mass produced non-silicon based chips?
In any case, it is claimed that Huawei or a Chinese research institute has already filed patents for a breakthrough in lithography.
The question is whether that can be brought to market as a product at some point.
I'd say a definite 'yes' with so much funding being ploughed into the Chinese industry and again, in well under twenty years.
I would take that bet. EUV didn't obsolete older lithography methods, they are still used for upper metal layers etc. If something better than EUV is eventually developed it is likely EUV will still be needed because whatever replaces it will be even more insanely expensive and difficult to build than the $100 million EUV scanners.
A breakthrough is always possible, but the combined might of the US, Europe, South Korea and Japan tried to find alternatives to EUV for 25 years when it became clear that the "obvious next step" of EUV was going to be far more difficult than first believed, and fantastically expensive to develop with no guarantee it will ever become production worthy. Many other avenues were explored, and all failed. The existing 193nm DUV technology was enhanced again and again with technology like immersion to work around limitations of standard lenses, multipatterning, and other tricks to make it last 20 years beyond the date at which roadmaps originally had it needing to be replaced by EUV. Eventually there were no alternatives remaining to improve DUV, and just in the nick of time they finally were able to make EUV work well enough to allow further progress in smaller and smaller feature sizes on chips.
So sure, China expending all their research might could make a breakthrough that eluded the best minds in the entire rest of the world funded by countless billions of dollars for 25 years. But I doubt it.
There are no unintended consequences here.
Ironically, it was SIA which wrote to the White House on several occasions warning of the damage a ban on Huawei would cause to its members.
Up to the first year of sanctions, Huawei was pumping billions into US semi conductor firms which, in turn, used those revenues to fund R&D for future advances.
So, losing revenues was bad, but what would be worse was 'de-Americanisation' by foreign firms which were suddenly seeing their own products being affected by the 'weaponisation' of US technology. That process is already underway and will also impact US interests.
By the way, SIA was not alone in warning the US to tread carefully with any sanctions. ASML has been very vocal and recently a high ranking South Korean official went on record as saying it shouldn't be put between a rock and a hard place as a result of US unilateral sanctions.
On top of that it was logical that Huawei, and all other affected Chinese firms, would have to 'brew their own' to be able to move forward. They have no alternative.
Huawei was clear on this and said it would take between two and three years to get US technology out of its systems.
They have already erradicated US technology from over 13,000 components.
They also invested in over 40 companies specialising in specific semi conductor related companies.
They upped R&D too. Patent filings a plenty.
One American EDA company executive even said he would rather Huawei used their tools by cracking them than see Huawei create competing tools.
Well, in both software and hardware and different platforms and industries, Huawei is going self sufficient and literally everyone (with perhaps the exception of Trump & Co) saw it coming.
EDA, chip packaging, photoresist, lithography...
The only difference is that it's now coming far faster than it otherwise would have - because of sanctions!
On a wider note China (as a whole) now has 7nm capacity. It might be low yield and more expensive but the US doesn't have any native 7nm capacity to call it's own. And if China were to refocus all its capacity (it won't) on 7nm, it's capacity would outstrip that of TSMC and Samsung - combined!
They went from 14nm to 7nm in two years,too.
SIA probably had a collective bed wetting episode after TechInsights published their report.
Of course the vast majority of chips used worldwide are not even produced on nodes anywhere near the cutting edge so China will be happy to improve yields and overall capacity on older nodes in the short term.
But we all know where they're heading.
As for Huawei in particular, my guess is that before year end they will announce a chip stacked solution on a flagship-like phone.
That will be another major step for them.
Unrelated or not, Qualcomm has recently said they expect no further 'material revenue' from Huawei.
I suspect something is in the pipe.
I've said before in several posts that all sanctions did was upset the build/buy balance. Back in the Good Old Days of global trade it just wasn't worth Huawei's while to invest in technologies that it could source, it put its development budget elsewhere. Once sanctions kicked in the build/buy tradeoff changed and the result was that all these technologies -- which aren't exclusively "American" anyway -- became not just cost effective to bring into China, it became a national necessity.
So we hold them up for a couple of years. All it costs us is a major market and the prospect of a formidable competitor going forward.
To my knowledge, China has always offered incentives only to its expatriate Chinese community in the West. This is what these media reports don't reveal. China is not recruiting engineers from India or engineers with European backgrounds. It focuses on recruiting Chinese engineers who left China as students and work in research and technology fields. It also recruits from Taiwan.
Huawei has over 20 Research Centres worldwide.
Mathematics research in France, Turkey and Russia (that last one is moving out of Russia at some point due to the geopolitical climate)
Imaging science in Finland.
RF in Canada.
Etc.
Obviously within China it's mostly Chinese but not limited to Chinese.
Huawei has been on a worldwide recruiting spree since sanctions kicked in. If you are truly talented in your field you can expect to earn around double the industry salary.
"Pope rumored to be an expert on Catholicism"
"Bears are now confirmed to defecate in the woods"
Seriously, what on Earth did we think was going to happen? The original report that I read was couched in terms of telling us how devious these Chinese are for circumventing our restrictions and, furthermore, failing to notify us they were doing so. It kind of beggars the imagination that anyone thought this wouldn't happen (and I suspect that's only the tip of the iceberg -- I've already read complaints about how the Chinese aren't publishing a lot of business and technical information any more so we can't easily figure out what they're doing.)
It is not unheard of, remember when SUN recruited a team from the Russian Academy of sciences to build the servers for the world wide web and designed the SPARK processors or when Vladimir Pentkovski was sent with a team from the Russian Academy of Sciences to teach Intel how to make processors starting with the Pentium = Pentkovsky, ... and teach them multitasking, multi-treading, out of order execution and many other modern processors tricks... The know how is all over the place and China will come out ahead regardless..... The Russians like parallel processing since 40 years ago and favor long instruction set programming and can do the same same job with half the processor speed, heat and node processes... There is a lot of advanced untapped processes in the world that are yet to show up....
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2025
