South Korea's $1.34 Billion Support For Semiconductors

In the next 10 years, the Korean government plans to invest 1.5 trillion won (about 1.34 billion US dollars) in semiconductors, research and development of next-generation memory chip materials, seek mutual symbiosis between IC design plants and fabs, and seek to become a global semiconductor company's production base.

In the post-Moore's Law era, the competitive situation will change rapidly. Major semiconductor countries such as the United States, Japan, and South Korea have jointly developed a response strategy based on national strength and industry.

A few days ago, South Korean Minister of Industry Baiyun announced that the South Korean government plans to invest 1.5 trillion won (about 1.34 billion US dollars) in semiconductor investment in the next 10 years, and will support the development of Korean semiconductors in three aspects, respectively, to develop next-generation memory chip materials and seek ICs. The design factory and the fab reciprocally symbiosis, and sought South Korea to become the production base of global semiconductor companies.

In my speech, there are only two spindles: first, to maintain the hegemony of the memory, 3D NAND continues to move forward; DRAM is fast moving, and new memory, such as MRAM, PCRAM, ReRAM, etc., needs new materials. The other is to eat into the foundry. In the past, the Korean fabs produced bulk commodities. The design company didn't have a place to raise the film. Now it is necessary to make both OEM and design. Memory plus foundry and design, probably a thousand years of the world, unified rivers and lakes.

DARPA in the United States has also been in the meeting, focusing on innovative ideas designed to provide new economic value for semiconductors. The semiconductor manufacturing power in the United States is somewhat lacking, but there is still an absolute advantage in the field of design. Another highlight is 3D monolithic stacking, like 3D NAND Flash, which is a vertical stack of simple repeating cell structures that requires only an additional 4 or 5 layers of reticle. The more complex lines, such as CPU plus memory, can't be integrated in a similar 3D NAND Flash process. 3D monolithic stacking makes the upper force faster than the 3D package and consumes less energy.

In the first half of this year, Japan announced the main direction of MRAM, silicon photonics and quantum computing. MRAM has been regarded by the Japanese government as a weapon to revitalize the Japanese semiconductor industry in the early years. Silicon photons and quantum computing are becoming more and more mature with the technology. There are several advantages in these directions: First, there is an urgent need to use advanced technology, so there is no need to spend a lot of money to catch up with the process immediately, and the evolution of the process is fast. Second, the future application market will expand rapidly. MRAM becomes an indispensable embedded memory in logic circuits. Silicon photons are crucial in the future of mass data transmission, and quantum computing is promising in pharmaceutical, chemical, and materials applications. Third, these components require new materials, and the materials industry is the traditional strength of Japan.

Comparing the national semiconductor industry development strategies of the United States, Japan and South Korea, I think the United States and Japan are slightly better. South Korea's strategy focuses on the development and expansion of existing industrial structures. The prospects of the United States and Japan are relatively broad, and it is expected to be a long-term industry change.

This is the national development strategy that the major semiconductor countries foresee that the competitive situation in the era of Moore's Law will change rapidly and that the country's strength and industry jointly develop.