It is not that once China overcomes the chip problem, the chip becomes the price of cabbage; but the chip has become the price of cabbage, hindering China from conquering the chip problem; the current status of the entire chip industry is the price of cabbage; we often see that a foreign device, or It is a shield machine, or a certain device, which China has made only one-third of foreign products. It will quickly defeat foreign products, and Chinese products will become the protagonist of the market; this is definitely not the case in the chip industry; a chip, if you do not consider research and development Cost, the cost of the chip is three 1: wafer cost (proportional to the chip DIE area), 2: packaging cost (determined by the packaging material and substrate size); 3: test cost (proportional to the test time); in other words, one The cost of a chip, if you remove the package, people in the industry can know about it; the cost of wafer is set by manufacturers such as TSMC and SMIC; the cost of packaging is set by ASE and Changjiang; this is the Matthew effect. The stronger the bargaining power of TSMC and ASE; the normal gross profit of a chip is around 30%-100% (you can see the financial reports of various listed companies, the average price is not a huge profit at all, and the profiteering industries are in prison. Closed); In fact, the current chip is the price of cabbage; for example, the development of a mobile phone SOC chip with an area of about 120 mm2 (7nm); packaging, testing, and yield are all included. This chip is not considered as the cost of research and development. A rough estimate is 100-150 yuan; 7nm wafer is more expensive; and the development cost of this chip is calculated according to 300 million yuan, (ip 100 million, MASK 50 million, manpower 100 million, other 50 million) If you sell 10 million The average cost of each chip is 30 yuan; this chip is 130-180 yuan; if it is sold for 300, it is estimated to be the price of cabbage; in addition, the price is reduced every year, and the life cycle of mobile phone SOC is very short; Qualcomm dares to buy 500, but also Add taxes; Qualcomm is a more profitable company, but it is not as profitable as expected; but compared to the investment, the chip is still the price of cabbage; the goal of the chip is to be more likely to be cheaper, so that more people can afford it. , Can be used, so that the research and development costs can be flat; if this chip only bought 1 million (it is also possible), each chip will be allocated 300 yuan, from the cost point of view, it will be 500 yuan, this The project must be at a loss; this is the difference between the chip industry and other industries; Intel is expensive, but who is right? If Alibaba Cloud purchases their processors or Huawei Cloud purchases them, what discount do you think they can get? The lowest price on the entire network; for these big customers, intel can make a little bit more, and it is definitely not a huge profit. This is the scale effect; this is the pain point of this industry. In the cabbage price chip industry, we are looking for opportunities. The hardest thing is; therefore, to contend, it is to twist the water in the towel; for example, domestic manufacturers sell the same MCU, and the price of STMicroelectronics is the ceiling. It cannot be more expensive than ST, but it can only be cheaper under the same function; By analogy, for the same SSD controller, domestic manufacturers cannot be more expensive than Marvell; for example, HiSilicon’s IPC SOC camera chip is cheaper than Ambarella and TI’s, which is basically 40% off foreign chips, and finally won 60%. In the global market, TI completely gave up; the industry-recognized HiSilicon still has such a method, besides other small factories? Unfortunately, before the large-scale growth of domestic chips, there are also a group of companies that are very good at this routine. They have already squeezed European and American chip companies through this method; they are the chip design heroes of Taiwan: MediaTek, Realtek, Novatek, Neto Da, Phison, Hui Rong, etc., are low in price and high in quality, and the wafer is cheap (Guess why? They all have a good relationship with TSMC, and they also ship a lot). It is the “old driver” of the chip “cost-effective” industry; for example, in the mid-end chips (a few to dozens of fields) SSD controller chips, TWS headset chips, set-top box chips, smart speaker chips, WIFI chips, and even mobile phone chips And so on, all occupy the bulk of the shipment; the amazing gross profit achieved by European and American companies through technical means requires high technical barriers (this is after all, European and American engineers, big houses, do not work overtime, and take a few months of vacation a year. Confidence); Unlike Europe and the United States, the chip prices of Taiwanese companies make you regret entering this business. For a certain chip before the stock was out of stock, a Taiwanese factory bought 5 yuan, opened the package, and the cost was at least 3 yuan. As a latecomer, it was not easy to defeat the Taiwanese company through a price war in this industry, and they did not leave it for themselves. Too much profit, those latecomers, manufacturing, packaging, and testing costs are definitely higher than those of Taiwanese companies; Chinese chip design companies break through and encounter wars. First, China Taiwan, Japan, South Korea and Europe, and finally the United States; The mid-to-low-end chip encounters with the above-mentioned companies are “decisive victory and defeat, but also life and death”; if these Taiwanese companies are a group of impressive “sacred mountains”; then they need “foolish men to move mountains”; South Korea likes countercyclical investment Samsung, Samsung likes to expand production when the industry is low, especially NAND and DRAM. That is, everyone is losing money, but we have to increase production. In the past two years, the demand has increased. Everyone can earn some money to catch up with the bad years, and the whole industry loses money. In the past few years, the big names in the storage industry meant that when the industry was at a loss, the proportion of losses should not be higher than that of Samsung in the industry. Win with the same loss. In other words, the big guys are prepared for Samsung’s idea of defeating competitors. This competition is worse than the price of cabbage. None of these East Asian competitors is easy to engage in.