Semiconductor giant mergers and acquisitions speed up, large silicon wafer oligopoly intensified, domestic substitution urgently need to break through

Silicon is the production of chips, discrete devices, sensors and other semiconductor products, key materials, more than 90% of the current semiconductor products are manufactured using silicon-based materials, silicon accounted for the proportion of semiconductor materials market size of about 37%, ranked first in the semiconductor three core materials, so semiconductor wafers are known as the semiconductor industry, "food! Therefore, semiconductor wafers are known as the semiconductor industry's "food", although the total global market size is not large, but vital.

In the past 20 years, semiconductor silicon wafers have long been monopolized by a handful of oligopolies such as Japan's Shin-Etsu, Japan's Katsushika (SUMCO), Universal Wafers, Germany's Siltronic, and South Korea's SK Siltron, etc. In 2019, the industry's top five companies accounted for as much as 92% of the proportion of global semiconductor wafer industry sales, and more than 90% of China's demand for wafers relies on imports, basically no domestic production, at present, the silicon wafer monopoly situation is still intensifying.

November 30, the German wafer maker Siltronic AG said, is carrying out in-depth negotiations with the global wafer, the latter intends to 3.75 billion euros (about $ 4.5 billion) will be acquired, the two sides are expected to be in the second week of December, to obtain the approval of the supervisory board of Siltronic and the board of directors of the global wafers, after the signing of the BCA.

Universal Crystal Chairman Xu Xiulan pointed out that both sides believe that the combination of business will be very effective, will be more complementary to the effective investment, and thus expand production capacity.

Before the merger, Germany Siltronic for the world's fourth-largest silicon wafer fab, market share of about 7%, the global crystal for the world's third-largest factory, market share of 18%, after the completion of the acquisition, the global crystal in the world's silicon wafer market share is expected to jump to 25%, close to Japan's 28% of the Katsumoto, and that means that the global wafer market will be presented to the Japanese Shin-Etsu, Katsumoto, global crystal, SK Siltron, further monopolizing the market.

According to IC Insights report, this year, only ADI acquisition of Maxim, NVIDIA acquisition of ARM, SK Hynix acquisition of Intel's storage business, AMD's acquisition of Xilinx four acquisitions of the transaction amounted to as much as 105 billion U.S. dollars, plus $ 10 billion acquisition of Inphi by Marvell, Universal Crystal $ 4.5 billion acquisition of German Siltronic, 2020 semiconductor mergers and acquisitions have hit a record high.

Last year, South Korea's SK Siltron to prevent Japan's export restrictions, the acquisition of DuPont Silicon Carbide Wafer Division, in the current rise of regional globalization, the giants embraced the warmth of their respective strengths through mergers and acquisitions to strengthen their respective advantages to cope with rapid change and complexity of the situation.

Taking semiconductor equipment giant Applied Materials as an example, in the decades prior to 2018, Applied Materials has long sat in the position of the world's first supplier of semiconductor equipment, by virtue of a comprehensive and powerful product line, especially in the semiconductor manufacturing front-channel equipment with a high level of technical content, the company has a fairly deep technological expertise.

But historically, Applied Materials has strengthened itself through a series of mergers and acquisitions (M&A). Although there was only one core business-related M&A during the 30-year period from 1967-1996, Applied Materials initiated 14 M&As during the 1997-2007 decade to continually refine its product portfolio.

As of today, Applied Materials' product line covers dozens of devices for semiconductor manufacturing, including atomic deposition, chemical vapor deposition, physical vapor deposition, ion implanter, etcher, chemical polishing, and wafer inspection equipment, etc. It is expected that Applied Materials' semiconductor equipment market share will increase from last year's 15.9% to 18.8% in 2020.

Checked the domestic semiconductor overseas mergers and acquisitions in recent years, there are mainly five:

2013 ZiGuang Group 1.78 billion U.S. dollars to buy Spreadtrum, 2014 907 million U.S. dollars to buy Ruidiqo, then merged to become ZiGuang Spreadtrum; 2015 Hefei Ruicheng 1.8 billion U.S. dollars to buy high-performance RF power amplifier manufacturers AMPLEON; 2016 CITIC Capital, Beijing Chingxin Huachuang Investment and Jinshi Investment $1.9 billion acquisition of CMOS sensor manufacturer Howell Technology;

2016 Changdian Technology to $780 million acquisition of Singaporean packaging and testing plant Jinko Xingpeng; in 2017, Jianguang Asset $2.75 billion acquisition of NXP's standard parts business, and this year's June Ansell Semiconductor officially injected into Wintek Technology.

Other overseas acquisitions are basically below $500 million, especially the acceleration of domestic substitution in the last three years, there are few overseas mergers and acquisitions of large cases, thinking about the domestic semiconductor plant is not bad money, but why still can not buy?

In fact, domestic enterprises overseas mergers and acquisitions, especially semiconductor overseas mergers and acquisitions is really not a problem that can be solved with more money, as early as 1996, the West 42 countries signed a group of restrictions on export control mechanism - "Wassenaar Agreement", in short, that is, the members of the domestic technology transfer or export without reporting, but the transfer to non-member countries need to be reported. The transfer to non-member countries need to be reported, so as to achieve the purpose of technology transfer regulation and control.

And this agreement is very up-to-date, to large silicon wafers, for example, the revised Wassenaar Agreement at the end of 2019, there is a new article on the export control of 12-inch large silicon wafer technology, pointing directly to the Chinese IC 14-nanometer process process, as well as upstream large silicon wafers applicable to 14-nanometer process.

Blockade is not only pull high-purity monocrystalline silicon ingot equipment and materials, but also from the cutting and polishing of silicon wafers on the specific parameters of the restrictions, specifically for the 14-nanometer process process applicable to a variety of silicon wafers. Therefore, small to specific product parameters can be arranged clearly, not to mention direct mergers and acquisitions of advanced enterprises, basically impossible.

If the United States unilaterally suppress Huawei is by virtue of its own strength, then through the Wassenaar Agreement to limit technology exports, is equivalent to pulling the circle around the world, limiting the circle of technology exports, outside the circle want to use, can only be used to cheap labor in exchange for the output of high value-added finished products in the circle, the Wassenaar Agreement is equivalent to the "Golden Bell The Wassenaar Agreement is equivalent to the "Golden Bell Shield". The Wassenaar Agreement is the equivalent of a "golden bell". So the mergers and acquisitions do not come to the company, only to build their own.

In fact, silicon wafers for photovoltaics have been played by domestic manufacturers out of the cabbage price, semiconductor wafers used by foreign monopolies, the difficulty lies in the purity of monocrystalline silicon and the control of internal defects, we do not do well.

In terms of purity, photovoltaic silicon wafers with 6 9 is enough, but semiconductor wafers need 11 9, that is, 99.9999999999999%, the problem is in the purity of this above, pull out of the monocrystalline silicon ingot purity is not enough, internal defects, stress, warping degree with foreign countries there is a gap between the chip made out of the yield is relatively low.

So in order to yield, foundries are willing to spend a high price to buy higher-quality wafers, rather than spend a low price to buy low-quality wafers, because it will lead to the final chip yield, China's production of silicon wafers can not be opened to the international market is the evidence.

The purification of monocrystalline silicon and crystal defect control, based on the accumulation of long-term practical experience and field error summary, which is highly confidential technology of various manufacturers, because of the factors in this regard, foreign wafer factories have not set up factories in the country.

At present, Shanghai silicon industry to break China's 12-inch (300mm) semiconductor wafer localization rate of almost 0 situation, to promote China's semiconductor key materials production technology independent control of the process; in the Central shares now also has 3-12 inches of full-size semiconductor wafer products mass production capacity.

Overall, the mainstream products in the global market are 12 inches, with a usage ratio of more than 70%, which are mainly used in high-end chips for smartphones, computers, artificial intelligence, and solid-state hard disks, etc. At present, 4-6 inches of wafers are used in the production of semiconductors. At present, 4-6 inches of silicon has been able to meet domestic demand, 8 inches is also maturing, entering a large-scale domestic replacement stage, but 12 inches has just entered the primary stage, is still faced with EPI, bit error and many other problems to be solved, the road to replacement is still a long way to go.