China’s silicon-self-sufficiency plan likely to miss targets due to Factories Not Present error
- Reference: 1591671911
- News link: https://www.theregister.co.uk/2020/06/09/china_semiconductor_manufacturing_targets_missed/
- Source link:
China wants to make its own silicon because it realises it is exposed to trade disputes that can – as has come to pass with the USA’s ban on Huawei – make it hard for local companies to source the components they need. Those fears, and a desire to grow local industry to achieve global leadership positions, led the country to prioritise development of its semiconductor industry in the 2015 policy called “Made in China 2025”.
But IC insights suggests the semiconductor part of that plan will struggle to achieve a target of 70 percent local production by 2025, instead reaching around 21 percent in 2024.
One reason for the miss is that plenty of non-Chinese semiconductor manufacturers operate in the Middle Kingdom.
“Of the $19.5 billion worth of ICs manufactured in China last year, China-headquartered companies produced only $7.6 billion (38.7%), accounting for only 6.1% of the country’s $124.6 billion IC market. TSMC, SK Hynix, Samsung, Intel, and other foreign companies that have IC wafer fabs located in China produced the rest,” the analysts write.
In some fields, China just isn’t anywhere near up to speed. IC Insights used DRAM as an example of that predicament:
“Consider that China’s first indigenous DRAM supplier, Changxin Memory Technologies (CXMT), only began limited production of its first DRAM products in 4Q19. This company has a few thousand employees and a capital spending budget of about $1.5 billion per year. In contrast, Micron and SK Hynix each have well over 30,000 employees and Samsung’s memory division is estimated to have over 40,000. Moreover, in 2019, the combined capital spending from Samsung, SK Hynix, and Micron was $39.7 billion. Now that’s a reality check!”
“Currently, China is putting on a brave face with regard to its future IC industry capabilities,” the analysts conclude, adding that its team “believes it is essentially impossible for China to make significant strides in becoming self-sufficient for its IC needs (memory and non-memory) within the next 5 years and probably not even within the next 10 years.”
Which gives the likes of Intel and Samsung ten years to figure out what happens once the China market disappears. And that worrying can start against a background of what IC Insights [1]predicts will be an industry-wide COVID-19-caused Q2 semiconductor sales dip of negative five percent.
Some semiconductor firms have predicted growth, principally from 5G and flash memory. But the rest of the industry sees some pain. ®
Sponsored: [2]Running Your Modern .NET Application on Kubernetes
[1] https://www.icinsights.com/news/bulletins/Sampling-Of-2Q-Semiconductor-Sales-Guidance-Now-At-5/
[2] https://go.theregister.com/tl/1956/-8477/running-your-modern-net-application-on-kubernetes?td=wptl1956
"Of the $19.5 billion worth of ICs manufactured in China last year, China-headquartered companies produced only $7.6 billion (38.7%), accounting for only 6.1%"
I'd be interested to know why that is. Are their own chips too low quality? Too power intensive? Too expensive? Adoption barriers to high? Production output too slow? Even if the government were to offer subsidies, would be overall disadvantages still be too high?
" In contrast, Micron and SK Hynix each have well over 30,000 employees and Samsung’s memory division is estimated to have over 40,000. "
That's fools logic. They can scale up quite easily, this isn't a cottage industry here with individual weavers turning out individual carpets. More people is worse not better, it leads to inertia. The machines they need, can be second sourced in Germany and UK (and some in China), so the factory kit is replaceable. The software design tools they already have. None of that is irreplaceable.
As to whether they can do it in 4 1/2 years? That seems a trivial goal. They could do that in 1 year if you pushed them.
Look at the result of the tariffs. How many Chinese made items have had their manufacturing moved to the USA? ~none. Close to zero. You still cannot make the products in the USA, because they could not be exported at a competitive price to markets where US tariffs don't exist to skew the market in USA favor. I doubt they even have the technology needed at this point to make stuff, even if it was only for US markets.
It's not going to change either, USA trades at a loss and has since the 90s. It's propped up by money printing. Each dollar made is leveraged up, and given to the Mnuchins/DeVos's/Carlson's/Swanson's/Beal's. That inflates the GDP number. It's why housing is overpriced, stocks and other assets are overpriced, its simply billionaires are awash with free money and have nowhere productive to put it. So they pump up asset values for the rest of Americans.
The whole economy is locked into that. To borrow money to fund that increasing deficit, that GDP number needs to keep increasing, to keep increasing as the real economy underneath shrinks, they have to print more and more money. Productive workers cannot work for low wages, if their houses costs so much, and their pension plan is built on overpriced stocks, and their health insurance costs an insane amount. So wages can never be competitive.
Contrast that with the export tariffs. Soy, Maize, Wheat, Chinese markets poof gone. Farm exports easily replaced. Huawei simply re-sourced US chips and replaced Google store with their own.
Trump's Atlantic city projects. $700 million in junk bonds supported by $820 a day in profits. That's what it reminds me of. All bluster, smoke and mirrors.
Only a decade
"Which gives the likes of Intel and Samsung ten years to figure out what happens once the China market disappears."
And roughly eleven years to figure out how to deal with Chinese competition in non-Chinese markets. Less than that really because semiconductor fabs are, I'm told, rather specialized, so China will presumably reach 100% plus domestic capacity in some types of semiconductor before the decade is out and will, one assumes, start selling their excess product into overseas markets.