SK Hynix and Intel jointly announced the signing of an acquisition agreement. SK Hynix will acquire Intel’s NAND flash memory and storage business for US$9 billion in all cash. Both parties expect to obtain the required government agency’s license by the end of 2021, and in 2025 The final delivery was completed in March.
According to the announcement issued by SK Hynix, the NAND flash memory and storage business acquisition agreement signed by both parties includes: NAND SSD (solid state drive business), NAND flash memory and wafer business, and Intel's NAND flash memory manufacturing plant in Dalian, China.
As Intel's leading storage technology-Optane (Ao Teng) business will not be included in the acquisition agreement, but will continue to be retained by Intel.
The entire acquisition process will be divided into two phases: the first phase will start after obtaining government permission. SK Hynix will pay USD 7 billion to acquire the NAND SSD business from Intel, including related intellectual property rights and employees, as well as the Dalian NAND flash memory manufacturing plant ; The second phase will be carried out at the final delivery in March 2025. SK Hynix will pay the final payment of US$2 billion to take over other related assets of Intel, including intellectual property rights related to the production and design of NAND flash memory wafers, R&D personnel and Dalian NAND flash memory Employees of the manufacturing plant.
Intel: Get rid of the shackles and go lightly
Intel’s sale of NAND flash memory and storage business did not actually surprise the industry. It first parted ways with Micron, completely sold its IMFT factory shares, and then gave up the baseband chip business and completely withdrew from the main smartphone-related business. In the past two years, competition Intel's gradual decline in power seems to have only one theme-to reduce the burden and focus on the core.
As the king of the semiconductor field in the PC era, after experiencing a round of strategic mistakes, Intel has not only been slightly fatigued in the mobile Internet era, even its proud main battlefield has also been constantly challenged by industry upstarts, and finally had to A series of mistakes such as the choice of technical direction and execution ability pay for the bill. Under both internal and external pressures, Intel not only gradually lost its original competitive advantage, but also had to abandon the "self-produced and self-sold" model that had been operating for many years, and for the first time this year expressed its willingness to outsource orders.
In the NAND market, Intel has, as always, shown a state of lack of competition and involuntary entanglement. This is because the price of NAND flash memory is greatly affected by market supply and demand, and the R&D and production of flash memory requires a large amount of capital investment. It is difficult to guarantee profits, and selling related businesses is undoubtedly an option to reduce risks.
According to data, Intel’s Dalian factory spent 2.5 billion US dollars to build it after 3 years, and is fully responsible for Intel’s NAND flash memory business. However, Intel, which is burdened by huge manufacturing pressures and has spent tremendous efforts to perfect a complete chip semiconductor industry chain, has found that the return from huge investment is slow and very limited.
● From a technical perspective, Intel is different from other manufacturers. It is currently the only company that adheres to the floating gate structure (Floating Gate) as a manufacturer of 3D NAND Flash, and is unique among the manufacturers that adopt the charge trap structure. ；
● From the perspective of production, Intel only operates a large NAND factory in the Dalian factory in China, which accounts for about 6%-7% of global NAND production;
● From the perspective of market share, Samsung firmly controls the NAND market and occupies a dominant position. Among the six companies dominated by Samsung, Kioxia, Western Digital, Micron, SK Hynix and Intel, Intel has only 11% of the market. , At the bottom. And among these six companies, except Intel, other companies regard storage business as their main business or main profit point. Only Intel's storage business is not the main business, and its revenue share is not high.
● From the perspective of revenue, Intel's market revenue in this area is also not stable. According to the financial report, Intel's non-variable storage solutions business has been in a continuous loss state since 2016. From 2016 to 2018, it lost 540 million U.S. dollars, 260 million U.S. dollars, and 5 million U.S. dollars. It didn't make a profit of 120 million U.S. dollars until 2019 .
Coupled with the eagerness for success and the pursuit of high-profit goals, it seems not surprising that Intel sells NAND-related businesses. According to related reports, as early as July 2019, there was news that SK Hynix planned to acquire Intel's entire Dalian factory and 3D NAND business, and the news also claimed that Intel would only continue to retain XPoint-related memory storage technology.
There is no doubt that the sale of NAND flash memory and storage business will enable Intel to focus more on its core business.
Intel CEO Bob Swan said in a statement, "This transaction will allow us to further prioritize investments in differentiated technologies." Intel will invest in the proceeds of this transaction, focusing on key businesses with long-term growth potential, including artificial intelligence, 5G networks, and edge devices related to autonomous driving.
SK Hynix: Trying the courage and seeking outburst
On the other hand, for SK Hynix, the acquisition of Intel's NAND flash memory and storage business is also traceable. As the world's second-largest memory chip manufacturer, although its market share of DRAM products is second only to Samsung Electronics, its market ranking for NAND is still relatively low.
Secondly, SK Hynix's internal revenue is extremely unbalanced. According to the financial report, its DRAM business accounted for 73% of SK Hynix's total revenue in the second quarter of 2020, while its NAND business was only 24%.
Through the acquisition of Intel's NAND business, SK Hynix is expected to increase its share of the NAND market to more than 20% in the short term, thereby narrowing the gap with Samsung Electronics, which is in a leading position. At the same time, SK Hynix will also get help in the "arms competition" with Samsung Electronics. It is reported that Samsung announced on August 28 this year that it will invest 7 billion US dollars in the future to expand the production of NAND chips in Xi’an Samsung Electronics. It is natural for SK Hynix to not be left behind. Therefore, it is very important for SK Hynix to obtain the Dalian factory.
In the long run, the industry consolidation caused by this transaction will also help reduce industry competition and help SK Hynix surpass other competitors. Both Intel and SK Hynix have the technology to develop and independently produce NAND flash memory, but their market positioning is different. The former is enterprise-level SSD and flash memory chips, while the main battlefield of the latter is on the consumer side.
However, the interdependence of pros and cons is an unchanging law. While obtaining huge benefits, SK Hynix also bears no small risks.
Cape Investment & Securities analyst Park Sung-soon said for the first time: "In the short term, this transaction will bring financial burden to SK Hynix. Due to oversupply, it is expected that the company will continue to announce its NAND business losses."
In addition, SK Hynix and Intel are also facing technical running-in problems. The two technologies mentioned above belong to the charge trapping camp and the floating gate structure camp. They are both in product storage density, performance and reliability. Each has its own advantages and disadvantages. Although there is not much difference in the manufacturing process and the required equipment materials, it is only the adjustment of the product structure and process parameters, but in the high-precision storage industry, any details have a huge impact.
Therefore, whether it is technological integration or seeking a balance between different product architectures in order to achieve the maximum synergy, it may be the focus that SK Hynix has to focus on.
In the era of great changes in the semiconductor field, industry integration continues, mergers and acquisitions become common practice, and the industry continues to optimize. The veteran giants urgently need to get rid of the shackles and go lightly, fortunately to focus more on the core areas to obtain long-term returns; the industry giants can't wait to arm themselves, strengthen the competitiveness of the industry, and challenge the dominant position of the industry. In the end, is it a win-win or two losses?