Array
(
    [content] => 
    [params] => Array
        (
            [0] => /forum/threads/korea-overtaken-by-china-in-semiconductor-technology-says-a-report.22160/
        )

    [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
)

'Korea overtaken by China in semiconductor technology,' says a report

Daniel Nenni

Admin
Staff member
67bba6d11c64d273823e.jpg


Korea has been overtaken by China in its core technological capabilities in the semiconductor sector in just two years, according to a new report. This is a warning sign that Korea's semiconductor industry is losing its technological momentum amid various challenges, including a government leadership vacuum and trade threats from U.S. President Donald Trump.

According to a report by the Korea Institute of Science & Technology Evaluation and Planning (KISTEP) on Sunday, a survey of 39 semiconductor experts in Korea found that, as of last year, Korea lagged behind China in four of the five core semiconductor technologies: high-density memory, high-performance artificial intelligence (AI) semiconductors, power semiconductors, and next-generation high-performance sensing. It was rated at the same level as China in advanced packaging technology. In the previous 2022 survey, Korea rated ahead of China in the memory, sensing, and packaging sectors but has since been overtaken or caught up with in just two years

The survey was conducted by selecting semiconductor experts from those who participated in KISTEP’s 2022 Technology Level Assessment. The assessment is conducted every two years to eval‎uate the country's core technologies in accordance with the Framework Act on Science and Technology. The semiconductor eval‎uators included university professors and researchers from industry, academia, and research institutes, as well as professionals from large companies and small- and medium-sized enterprises in materials, components, and equipment.

Experts are increasingly concerned that Korea has ceded its major strengths in memory, sensing, and packaging technologies to China. The report found that while Korea remains ahead of China in process and mass production technologies within the semiconductor technology lifecycle, it lags behind China in basic and fundamental research and design technologies and even ranks last among other major countries.

China's push into Korea's traditional major technologies is not limited to semiconductors. Last year, Chinese companies overtook Korea for the first time in global TV shipments, putting their super-sized products at the forefront. In the premium smartphone and display markets, Chinese companies are also threatening Samsung and LG's market share. "Securing core technologies that can serve as a strategic card for Korea in the global value chain is crucial," said Jeong Eu-jin, a researcher at KISTEP. "Efforts to secure talent are urgently needed, including incentives for the industry to nurture skilled individuals and immigration policies to attract overseas professionals."

 
A very strange take on the memory. Korean retreat from flash market is not really what it looks.

Both Hynix, and Samsung make tons of money on large server SSDs, and have lost interest in the consumer space. On the other hand, server SSD's are not particularly power, or performance sensitive, and they started to lag behind companies which still care about that.

Particularly for Samsung, which always was a volume player, walking away from commodity flash market was a move which left them with significantly overcapacity on old flash fabs.
 
Interesting report. China is not a leader yet. but if other countries continue to threaten not to trade with them, they will be the leader in all semis in 5 years.

China has smart people, great universities, hard working people, tons of investment capital, a billion customers.
 
Interesting report. China is not a leader yet. but if other countries continue to threaten not to trade with them, they will be the leader in all semis in 5 years.

Almost all electronic components are made outside of PRC, almost all electronics is made in PRC.

Either finished electronics manufacturing moves out of PRC, or components will get in, given that even GPUs not made in such large volumes are freely diverted to China.
 
I don't necessarily agree with the article but I hope everyone in the world knows that at some point in time it will happen. China will semiconductor pass us all if we don't change course. The US is making biggly changes and as a result so will the rest of the world. Will this help or hurt? Nobody knows but we will certainly find out. I stay optimistic.
 
Interesting report. China is not a leader yet. but if other countries continue to threaten not to trade with them, they will be the leader in all semis in 5 years.

China has smart people, great universities, hard working people, tons of investment capital, a billion customers.

China has the talent and the market, but success also requires creativity, which isn't just about understanding existing processes. While Chinese engineers are good at tweaking designs and making incremental improvements, semiconductor innovation requires bold thinking and strong overall planning - qualities I'm not sure are yet widespread in China's industry.
 
China has the talent and the market, but success also requires creativity, which isn't just about understanding existing processes. While Chinese engineers are good at tweaking designs and making incremental improvements, semiconductor innovation requires bold thinking and strong overall planning - qualities I'm not sure are yet widespread in China's industry.

It only takes one , however that one will need some very high level support as the stigma of failure is very very high in this part of the world.

Therefore small iterative improvements on established processes seems the preferred route.
 
China has the talent and the market, but success also requires creativity, which isn't just about understanding existing processes. While Chinese engineers are good at tweaking designs and making incremental improvements, semiconductor innovation requires bold thinking and strong overall planning - qualities I'm not sure are yet widespread in China's industry.

I think this was the case in China, imitation versus innovation, but things have changed. It really was a do or die situation with the US blockading semiconductors. It is like blockading oil or natural gas. It is a threat to modern life.
 
China is not a leader yet. but if other countries continue to threaten not to trade with them, they will be the leader in all semis in 5 years.
If this analysis is correct, we better not tell the Chinese. Otherwise they'll close all necessary trade from their side and get to leadership in 5 years.
 
A very strange take on the memory. Korean retreat from flash market is not really what it looks.

Both Hynix, and Samsung make tons of money on large server SSDs, and have lost interest in the consumer space. On the other hand, server SSD's are not particularly power, or performance sensitive, and they started to lag behind companies which still care about that.

Particularly for Samsung, which always was a volume player, walking away from commodity flash market was a move which left them with significantly overcapacity on old flash fabs.
Samsung rumored to adopt YMTC's patent for NAND, which isn't a good sign:
 
Samsung rumored to adopt YMTC's patent for NAND, which isn't a good sign:
Hybrid bonding is the NAND future and YMTC was the NAND leader on this. It is possible everyone pays royalties to YMTC. I believe Xperi had agreements with other NAND providers. Other companies are copying Chinese company YMTC.... interesting ;)
 
Samsung rumored to adopt YMTC's patent for NAND, which isn't a good sign:

Which us strange, given that Samsung was first to do serious R&D on wafer bonding, and they were ahead of the whole world with advanced packaging a decade ago (WIO).

PRC's wafer bonding tech, ironically, has its roots with LED blinkers, and cheapo SFP transmitter chip-lasers.
 
YMTC began with bonding to glue two NAND dies, doubling the layer count. They did it because they were far behind Koreans in the number of layers they can do with monolithic, and the layer number is highly symbolic in the industry.

So they glued two 96 layer dies, and got 192 layer in one package at an extra expense, and yield losses. But later, the started with adding a CMOS die, through the same process, which saved them.

One faulty column, or flash block is easily repairable, but a defect in driver electronics scraps the whole die. By separating the two, you make the money making NAND die way more repairable, and get huge yield boost.
 
Back
Top