Gexin, the world's second largest semiconductor foundry, announced that it will be indefinitely in the 7nm FinFET advanced process development. After the UMC, the semiconductor manufacturers competed for more advanced processes, and the outside world worried about the impact on the global foundry wafer industry, and analyzed for several aspects.
How will the wafer foundry market structure change?
Participated in the 7-nanometer process competition, including leading Taiwan Semiconductor Manufacturing Co., Ltd., Gexin, Samsung and Intel. Each factory is scheduled to provide corresponding technology for 7 nanometer nodes on behalf of industrialists in 2018, allowing designers to have more choices, which can be expanded in 7 The bargaining power of nano-designers.
The core 7nm R&D route has been finalized in 2016 and was originally expected to be mass-produced in 2018. However, as the pace of mass production approaches, although the major customers have decided to OEM partners, the reliability of these customers to the grid core and The approval of product schedules is still limited. Without the support of sufficient order volume, the continued development of the 7-nanometer process has no economic benefits for the grid core. With the 7-nanometer market structure already fixed, Gexin announced that the investment was interrupted and exited the competition. Its agreement with customers at this node has failed, and the impact on the global 7nm wafer foundry industry pattern is limited.
With the indefinite suspension of the grid, the market's focus on 7 nanometers may be passed on to TSMC, Intel and Samsung.
Why did Gexin decide to abandon the 7nm process development?
For the 7-nanometer production cycle and application development, the process is more complicated than 16/14nm. Therefore, as long as the EUV equipment can provide sufficient mass production, the major manufacturers will partially import EUV at least on the 7nm Gen2. From the process level, 7 nanometers is a conversion node for advanced exposure equipment.
In terms of the current product structure, today's integrated circuits are in maturity and communication under the two main categories, the main product sales are entering a mature stage, and the short-term application of 7nm or 5nm products is not enough to provide new process nodes. The amount of return required for R&D and commissioning will face significant challenges for the capital investment recovery of wafer foundries.
This is not only the main reason for UMC and Gexin to withdraw from the advanced process competition, but also the reason why leading manufacturers TSMC and Intel hope to win more market share in order to maintain sufficient growth momentum of the plant. If the overall demand of the existing process nodes is not rising enough, after the suspension of advanced process investment, the core and UMC will improve the overall financial situation, there will be more space in the pricing, and compete with the existing process nodes and still at 7 nm. The competition in the competition.
TSMC takes a short-term profit and benefits, reflecting economies of scale
For the withdrawal of the grid core, the focus of market competition is placed on the two major factories of TSMC and Samsung. According to the research report of Jibang Consulting, TSMC’s market share in the first half of 2018 is as high as 56%, accounting for more than half of the market share. It has a very poor factory. The market generally believes that the second-ranked Gexin will withdraw from this competition, which is more beneficial than TSMC.
In this regard, Jibang Consulting believes that in the short term, the above arguments are correct. From the perspective of free market, TSMC needs to demonstrate sufficient scale economy through the order of Computing, and the withdrawal of Gexin makes Supermicro advanced. Orders for the process are mostly transferred to TSMC this year and next, which will support it to move smoothly to the next node.
However, after TSMC successively won several major customers at 7 nanometers, it also caused greater pressure on competitors Samsung and Intel. These major customers are also bound to consider the risk of not having a second supplier. The customer's investment strategy after the effectiveness and risk will also influence the long-term development of TSMC. Therefore, whether TSMC has an advantage in this competition remains to be discussed.