Automotive Electronics

 In recent decades, the use of semiconductor products in automobiles has experienced rapid expansion, making it one of the fastest growing market segments. A single car requires dozens of different types of chips, each with multiple variations, totaling hundreds or even thousands of chips throughout the vehicle. From simple functions like door and window controls to critical operations such as engine ignition, automotive chips are indispensable. The demand for semiconductor chips is even higher in electric vehicles, especially those equipped with advanced autonomous driving capabilities reaching up to L4 or 3000 per car. Consequently, the semiconductor supply chain has become an essential component in manufacturers' product planning processes. When selecting components for automotive products, quality, reliability and service life are paramount considerations; thus establishing a trusted partnership is crucial. Although power semiconductors first appeared in automotive applications as early as the 1850s and simple microcontrollers were introduced in the 1970s, their usage has significantly evolved with advancements in safety features and the introduction of hybrid and pure electric vehicles.

In 2019-2021, the auto industry experienced a chip shortage, and many auto chips saw their prices skyrocket, which was due to industry reasons. Most auto chips are produced using the IDM vertically integrated model, where design, manufacturing, and packaging are all handled in-house, with only a few spillover orders being outsourced to foundries like TSMC. The top six players in MCU are all IDM mode; the top ten players in IGBT, except for Magna's transition to fabless, are all IDM mode. Under the IDM model, if companies A and B both produce similar chips, and A has a capacity of 1 million units, and B has a capacity of 300,000 units, then B's profit will not be 30% of A's, but rather likely to be in the red. Chips must be produced on a large scale to spread out research and development costs and capital expenditures, and sales volume directly determines the cost of the chip. Therefore, in standardized fields such as storage, panels, and auto chips, it is often the wealthy and powerful incumbents who initiate price wars, rather than new entrants. Without a certain market share, one does not even have the qualifications to engage in price wars. Therefore, compared to Qualcomm and NVIDIA's art of challenging the laws of physics on 12-inch wafers, the competitive landscape of auto chips is quite straightforward: controlling production. The construction of chip capacity takes a very long time, and during this process, the market supply and demand will undergo significant changes. If too much capacity is planned, there will be an oversupply; if too little capacity is planned, the market will be given to competitors. Before the chip shortage, the automotive chip market was stable, and the six major players had roughly estimated how much capacity they had. Even if chip prices fell beyond expectations, it was still within the prediction range. Even if the oversupply exceeded expectations, car manufacturers had limited choices of suppliers. At the very least, a mysterious fire could break out in the factory, which was not impossible. During this process, the six major players would also expand their product portfolios through large mergers and acquisitions, consolidate their market share advantage. Over the past decade, Japanese automotive chip giant RENESAS has acquired American companies Intersil and IDT, and British company Dialog, to fill in the gaps in power management chips, wireless network and data storage chips, and wireless communication. In such a competitive landscape, it is understandable what the chip shortage means for Chinese automotive chip companies. Due to the pessimistic expectations of the market for auto sales, car manufacturers have cut orders. Just then, Texas Instruments and Infineon (Infineon) began upgrading from 8-inch production lines to 12-inch production lines and closed old production lines. As a result, the market for new energy vehicles exploded, and chip prices soared to the sky. As long as there is a little capacity, it has become a scarce target, and the performance has soared. Renesas Electronics has recorded consecutive record-high net profits and revenue, while Texas Instruments (TI) has achieved double-digit revenue growth for seven consecutive quarters, and ON Semiconductor, a leading player in IGBT, has set a record for revenue.

Automotive chips have a lower technical content than military and aerospace chips, with more than 70% of automotive chips using mature processes of 90nm or above, and only 6% using advanced processes of 14nm or below. Therefore, domestic automotive chip companies seized the opportunity and achieved rapid development of domestic automotive chips. In just three years from 2020 to 2023, the number of Chinese automotive chip suppliers increased from dozens to more than 300, releasing a large amount of new capacity, followed by a price war. For car manufacturers, in the face of the inventory red line and soaring chip prices, vehicle compliance verification and product matrix are low priority issues. In order to ensure vehicle delivery, car manufacturers had no choice but to give domestic automotive chips the green light, replacing products that could not be supplied by overseas suppliers. On the other hand, many domestic new forces have rich engineering experience and are relatively familiar with their own electronic and electrical architectures, and are more inclined to use domestic chips. However, under the surface, many domestic automotive chips are still missing the final mile to go on-board - car compliance verification. Unlike consumer-grade chips such as smartphones, automotive chips need to consider more extreme working environments. For example, the temperature range required for the former is 0°C to +70°C, while the temperature range required for automotive chips is -40°C to +86°C. The validation cycle of the supply chain for car manufacturers is usually two to three years, and many products may have already reached the final testing and verification stage. Therefore, although the localization rate is just 10%, the Nikkei still exudes a strong sense of anxiety: even for highly technically challenging chips, China is expected to achieve domestic replacement within 5 to 10 years.