After more than half a year of demand adjustment, there is no shortage of general-purpose materials in the market, and there is even a surplus due to terminal cut orders. Many distributors are also feeling the pressure due to the reduction in demand. However, in the case of the reversal of the market wind, the expansion plan of the original IDM factory has not wavered, which is very different from the slashing of capital expenditure by wafer foundries and storage manufacturers.
So with the continuous influx of new materials into the market, is it necessary to say “farewell” to the lack of core? How to judge this issue will determine how distributors, terminal operators and other industrial chain links will deploy future strategies.
To meet long-term needs, IDM continues to expand counter-cyclically
There are many differences between the two mainstream operating models of IDM and foundry in the industry, and this is the case in terms of production expansion. The operation of wafer foundry is based on customer orders. When customers cut orders during a period of sluggish demand, it will naturally drive the foundry to slow down its production expansion. Otherwise, the expanded equipment and machines can only be idle, consuming depreciation costs. However, IDM is different from it. The production process is determined by itself, and counter-cyclical expansion is even a good strategy for advanced layout.
At the end of September, TI announced the initial production of its new 12-inch factory RFAB2 in Richardson, Texas, and will continue to expand production capacity in the next few months. In recent years, TI has been expanding its investment in 12-inch production capacity. Since 2017, more than half of TI’s analog revenue has come from 12-inch production lines. TI also planned to invest $30 billion last year to build four new 12-inch plants in Sherman, Texas. In May of this year, the first one broke ground and is expected to go into production in 2025.
Corresponding to TI, ST, which focuses on MCUs and takes into account analog devices, made a plan last year to double the overall 12-inch production capacity in Europe from 2020 to 2025, and selectively increase some non-12-inch production capacity. In addition to conventional devices, ST will also invest 730 million euros to build a silicon carbide (SiC) fab in Italy to meet the huge demand arising from the transition of fuel vehicles to new energy.
For ON Semiconductor, which can also enjoy a lot of dividends in the new energy market, now is also an opportunity to expand production that cannot be wasted. As the main original factory of power devices, ON Semiconductor will expand its 12-inch production capacity and deploy the silicon carbide industry in the next two years. For the latter, ON Semiconductor pointed out that the production capacity in the next five years will be 1.3 times that of the present.
In addition, NXP plans to invest US$2.6 billion to expand production in Austin, Texas in May this year; Infineon recently opened a new power semiconductor plant in Hungary; Microchip is considering investing US$3 billion to expand production in Oregon, US, and plans within a few years $5 billion to build a 12-inch fab. It can be seen that the enthusiasm of IDM to expand production has not slowed down due to the reversal of the chip market.
Supply and demand and policy orientation are superimposed, and there are various reasons for IDM to expand production
IDM’s production expansion plans were established during the most serious period of core shortage. Now the market situation has reversed and the shortage has not shaken the implementation of the original plan. The reasons for this come from many aspects.
First of all, even if the market as a whole is out of the lack of cores, there are still many high-end devices and automotive-grade materials that still have a large supply and demand gap. According to the information disclosed by the original factory, Infineon’s backlog of orders in the first three months of this year has increased from 31 billion euros in the fourth quarter of last year to 37 billion euros; ON Semiconductor has no longer received orders for IGBT products, and its production capacity this year and next year will also increase. All are booked. In terms of delivery time, Infineon’s IGBT delivery time is still 39-50 weeks, and the MOS tube material is about 50 weeks. The automotive MCUs of ST, NXP and other manufacturers are still in the distribution state, and the quotations are also maintained at a high level.
It is not only necessary to consider the immediate future, but with the continuous development of industries such as new energy/smart vehicles, industrial equipment and medical equipment, the “core content” of single equipment is also increasing, so the expansion of the original factory should also meet the long-term needs of the industry. Taking into account, even the increase in the number of cores used in a single device is more decisive than the increase in the overall sales of the device. For the original IDM factory, although the current market cycle has turned, the demand for the number of chips will not decline in the future, and it can be counter-cyclical layout at the moment.
In addition, recently, due to factors such as industrial security and geopolitical games, the United States and European countries have begun to subsidize the semiconductor industry to build factories and strengthen local production. Under the blessing of a large number of subsidies, semiconductor manufacturers have become more aggressive in expanding production, and even in order to win subsidies, they have to open more expansion projects and increase R&D efforts in order to seize more potential market share.
After going through the lack of core, will the market “grow rationally”?
The shortage of cores after the epidemic has caused the market to enter an “irrational” state. Now that demand has cooled, the original production capacity has been successfully opened, and the market situation is being reversed from both supply and demand. For a long time, there have been many claims that chip production capacity has always been able to cover demand, but it is only because of factors such as poor logistics and panic ordering that the apparent “chip shortage tide” is caused.
If the previous “lack of cores” is not a rigid shortage, then the current expansion of IDM will only make the supply of materials more and more abundant. However, oversupply should not be taken too seriously. With the continuous increase in the amount of chips in the fields of automobiles and industrial control, these expanded materials can find their destination. After the market returns to rationality, the opportunities that can be left to the majority of distributors and terminal operators will be easier to grasp.