One Week News | Auto semiconductor surprise cut single price..........

1. Car semiconductor shock transmission cut single price
The latest survey of foreign investment Morgan Stanley (Morgan Stanley) more disclosed that there is a downside risk of automotive semiconductor, in early March, some auto markets began to appear price competition, cut orders, price signs are obvious, confirmed that the auto market is no longer stable.
Auto manufacturers cut prices to the market to burn automotive semiconductor, including BMW, Geely and other indicator auto manufacturers recently for power management IC, gold oxygen half effect transistor (MOSFET), microcontroller (MCU) and other previously hot chip cut orders and asked suppliers to lower prices, affecting the former stock king Sily-KY, Hon Hai to invest in Fuding, Huaxin Lihua Group's Xintang and other Taiwan factories to receive orders.
Look at the automotive chip manufacturers, TI said in the earnings briefing, Q4 last year, except for the automotive business, all end markets were weak; NXP's automotive chip sales grew 25 per cent last year, while Renesas's automotive business grew nearly 40 per cent last year and both expect further growth in the current quarter. Infineon is even more blunt, the automotive division's capacity has been sold out in the 2023 fiscal year
2. HBM that bucked the trend in the cold winter of memory market
ChatGPT has moved from the downstream AI application to the upstream chip field. While pushing GPU and other AI chips to the peak, Chatgpt also greatly drives the market demand for the new generation of memory chip HBM (high bandwidth memory). Samsung and SK Hynix have seen a rapid increase in HBM orders and prices since the start of 2023. According to market sources, the price of HBM3 DRAM has recently increased by five times.
High Bandwidth Memory (HBM) is a new type of CPU/GPU memory chip. In fact, many DDR chips are stacked together and packaged together with GPU to achieve a large-capacity, high-bandwidth DDR composite array. Compared with traditional DRAM, HBM has significant advantages in data processing speed and performance, and is expected to gain widespread attention in the industry and be increasingly adopted.
3. When the market drops for the first time, how will the three domestic CIS enterprises go on?
In 2022, global CIS market sales reached $19 billion, down 7% from the previous year, marking the first market decline in 13 years.
In terms of segments, smart phones, which have been the biggest driving force for CIS market growth, accounted for less than 70% of the entire CIS market, showing a significant decline. At the same time, the market demand for surveillance and PC is also declining, while the share of automobile and industrial has increased significantly.
In general, CIS manufacturers still have the problem of overstocking. Low - and medium-end CIS inventory adjustment pressure is larger, auto structural growth point. As a result, major mobile phone manufacturers such as Sony, Samsung Electronics, and Howe are still adjusting to the pace of their customers, and will not recover until the third quarter at the earliest.
4. Lenovo and Nvidia work together to create a car computing 'Superbrain'
Instead of building cars, Lenovo aims to become an enabler of the "new Four modernizations" of automobiles. Lenovo was the first Tier 1 company to adopt NVIDIA's DRIVE Thor platform. In the future, the domain control platform architecture based on DRIVE Thor will become the high-end core product line of Lenovo car computing, and relevant products are expected to be mass-produced in early 2025.
DRIVE Thor can integrate various intelligent vehicle functions into the same AI computing platform, and realize intelligent driving and intelligent cockpit functions of the whole vehicle in a single computing platform, including automatic driving, parking, driver and passenger monitoring, digital dashboard and on-board infotainment system, etc.
5. The semiconductor industry is seeing the light of day as mobile-phone chips rise in price
It is a good sign that some mobile phone chips are already rising in price. According to Taiwan media reports, recently, mobile phone HD version driver and touch integration IC (TDDI) due to tight supply and demand unexpectedly price news, some IC design industry confirmed that the mainstream market HD picture quality TDDI chip price has risen by 10% since the beginning of March, this is the IC design factory after the second half of last year after the correction tide, the first time to see the price increase. The semiconductor industry is showing signs of prosperity.
Duntai said that the current order demand is still urgent, the price is determined by supply and demand. As Tianyu previously mentioned in the French Conference, mobile applications seem to have changed a little, and the trend of changing phones should slowly appear, and the related revenue will gradually recover. Lianyong is expected to meet the off-season this season, after the second quarter should be able to return to seasonal demand, the second half of the end demand is worth looking forward to. The short supply of HD version TDDI should continue to April, and the price may return to normal when production capacity supply returns to normal in May.
6. Is Lidar a tough road ahead or an uphill one?
How to choose between Lidar and pure vision?
Pure visual perception route selection takes machine vision as the core, and uses millimeter-wave radar + camera solution to realize automatic driving. Compared with the Lidar scheme, the pure visual perception route has lower cost, conforms to the logic of human eyes, and will produce qualitative changes after data accumulation reaches a certain scale, until it reaches the advantage of matching and surpassing the automatic driving performance of the Lidar scheme. It is a more advanced scheme in technology.
Tesla s FSD, Baidu s Apollo Lite, and Mobileye s Super Vision are the representatives
For the Lidar scheme, high-precision map + Lidar is chosen as the core sensor to realize automatic driving, which can help the vehicle realize rapid 3D modeling under the existing technical conditions, accurately restore the road condition information, and form a scheme that the computer can quickly identify, process and respond to. However, due to the high cost of lidar solutions and the high demand for chip computing power, it cannot be implemented in all vehicle classes at this stage. Representative enterprises are Xiaopeng, Nio, Baidu Apollo.
Three out of four preferred that lidar and vision fusion would become mainstream in the future.
A quarter of the population is firmly optimistic about the prospects of purely visual solutions. They are more concerned about the cost in the short term, the data limitations of lidar, and the leapfrog development of the vision technology itself.
According to the 2022 Automotive and Industrial Lidar Application Report released by Yole Intelligence, global shipments of lidar for automotive and industrial applications are expected to reach 300,000 units in 2021, with a market size of $2.1 billion, an increase of 18% compared to 2020. Among them, terrain mapping is still the largest application field of Lidar, accounting for 60% of the market share. It was followed by the industrial sector, with 27 per cent; Driverless taxis, ADAS (Advanced Driver assistance systems), wind energy and defense accounted for the remaining 13 percent.