【Lansheng Technology Information】STMicroelectronics STPM801 is the first to market the ideal diode controller with integrated thermal switching for automotive specifications, suitable for automotive functional safety applications.
This ideal diode controller from STMicroelectronics drives an external MOSFET switch, replacing the Schottky diodes commonly used in input reverse protection and output voltage hold-up circuits. The voltage drop across the MOSFET is lower than the forward voltage of the Schottky diode, therefore, the power dissipated during normal operation is also lower than that of the diode.‘
The ideal diode control circuit also provides an ORing controller for main power and backup battery power switching, ensuring uninterrupted power supply for safety-critical equipment such as autonomous driving and advanced driver assistance systems (ADAS).
A built-in hot-swap controller drives a second external N-channel MOSFET to protect the load during switching between main power and battery backup. The soft-start function charges a capacitor of known value connected to the gate with a constant current to control the turn-on transient event of the second MOSFET and avoid high inrush currents. The overvoltage and undervoltage pins cut off the output voltage as long as the input voltage is not within the specified threshold.
The STPM801 also has protection and monitoring functions and is suitable for systems with functional safety requirements up to ISO 26262 Automotive Safety Integrity Level (ASIL) Level D. These functions are integrated in a 5mm x 5mm VFQFN-32 package, saving PCB board area and minimizing external component count.
With an operating voltage range of 4V-65V, the STPM801 is tolerant to automotive electrical hazards, and its 25µA quiescent current minimizes battery drain when the vehicle is off. Target applications include zone/body ECUs, ADAS ECUs, high performance computing ECUs, infotainment ECUs, redundant power supply systems, and dual battery systems.
【Lansheng Technology Information】On August 9, according to the latest research report released by Ming-Chi Kuo, Qualcomm has stopped developing the Intel 20A chip.
He believes that Intel's lack of cooperation with first-line IC design companies such as Qualcomm will not be conducive to the growth of RibbonFET and PowerVia, and will further make Intel 18A R&D and mass production face higher uncertainties and risks.
Ming-Chi Kuo pointed out that after the advanced manufacturing process enters 7nm, the high-end orders of the first-line IC design industry are more important to the foundry.
According to public information, Intel plans to gradually move from Intel 10 to Intel 7 and Intel 4, and then continue to develop towards Intel 3, Intel 20A, and Intel 18A processes.
It is reported that Intel 7 has been mass-produced, Intel 4 will be launched in the second half of this year, and will be used for Core processors (Meter Lake), Intel 3 is advancing as planned, Intel 20A, Intel 18A test chips have been taped out, Intel 20A uses RibbonFET technology, which is similar to Samsung’s GAA transistor technology. In addition, Intel’s unique and industry’s first backside power transmission network will be used to optimize signal transmission by eliminating the need for power supply wiring on the front of the wafer. If all goes well If so, the Intel 20A process technology was originally planned to be mass-produced in 2024.
Although Intel announced that its CPU module will use the Intel 20A process when it announced Arrow Lake last year, there have been reports that it will switch to TSMC's N3 process. At present, it is reported that Intel has given up using the 20A process in Arrow Lake. All the chiplets above will be produced by TSMC.
【Lansheng Technology Information】Recently, major chip manufacturers have successively announced their Q2 financial reports for this year. Lansheng Technology has sorted out the Q2 revenue of TI, ST, NXP, AMD, Renesas and other chip manufacturers. Perhaps we can learn about the status of the chip industry. Current status and development prospects in the second half of the year.
1. TI's Q2 revenue in 2023 will be US$4.53 billion, an increase of 3% from the previous quarter, but a decrease of 13% from the same period last year, and a net profit of US$1.722 billion, a decrease of 25% from the same period last year. As TI's trump card business, analog chip revenue fell 18% to $3.2 billion from the previous quarter.
TI said that in addition to the automotive business, other end markets were weak, and customers continued to cut orders for new chips and instead relied on existing inventories. TI's own chip inventory has further increased, and the current inventory cycle has climbed to 207 days.
Looking forward to the next quarter, TI predicts that its revenue will be between US$4.36 billion and US$4.74 billion, which is lower than market expectations and indicates that the entire chip industry is still struggling.
2. ST's 2023 Q2 revenue will be US$4.53 billion, a year-on-year increase of 12.7%, and its net profit will be US$1 billion, a year-on-year increase of 15.5%. Both revenue and profit exceed market expectations.
The strong demand in the automotive and industrial markets has driven the overall performance of ST to continue to improve. ST's Automotive and Discrete Group (ADG) revenue rose 34% to $1.96 billion last quarter, accounting for 45% of overall revenue.
ST CEO said that ST's revenue performance continued to be driven by growth in the automotive and industrial segments, but was partially offset by a decline in personal electronics revenue.
3. NXP's Q2 revenue in 2023 is US$3.3 billion. Although the revenue fell slightly by 0.4% year-on-year, it exceeded market expectations of US$3.21 billion.
The automotive business has become the main revenue force of NXP. This quarter, the automotive business revenue reached US$1.866 billion, an increase of 2% month-on-month and 9% year-on-year, accounting for about 56.5% of the total revenue, more than "half of the country."
The CEO of NXP said that all of NXP's key terminal market revenue trends were better than expected, and the excellent performance made NXP believe that it is successfully surviving the cyclical downturn in consumer business.
In addition, due to the continued strength of the automotive, core industry and communication infrastructure businesses, NXP has also raised its Q3 revenue forecast. It is expected that Q3 revenue will be between 3.3 billion and 3.5 billion US dollars. It is understood that most of NXP's production capacity in the second half of the year is reserved by the demand in the automotive field, and the main growth point of NXP in the second half of the year is still automotive chips.
【Lansheng Technology Information】Samsung Electronics announced the research results of BSPDN (Backside Power Supply Network). This is the first time that Samsung Electronics has disclosed specific research results related to BSPDN.
BSPDN is a concept first presented at IMEC 2019. It is a design structure that improves power supply and signal lines and battery utilization bottlenecks by arranging power supply wiring on the backside of the wafer. Samsung Electronics said that by applying BSPDN, it succeeded in reducing the area by 14.8% compared with FSPDN (Front-End Power Supply Network). Specifically, in the two ARM circuits, the area was reduced by 10.6% and 19%, respectively. In addition, the wiring length was also reduced by 9.2%.
Current semiconductors are made using the FSPDN structure. They are arranged in the order of power line-signal line-transistor, but since the power line and signal line use the same resources, there will be problems such as bottlenecks. In addition, there is a disadvantage in that a large cost is consumed to expand the wiring layer according to the expansion of the transistor.
The semiconductor industry, including Samsung Electronics, has begun to focus on the BSPDN structure to overcome these structural limitations. Different from existing semiconductor structures, BSPDN is arranged in the order of signal lines-transistors-power lines. In the paper, Samsung Electronics explained that the utilization of the cells on the top of the wafer can be improved through BSPDN. In addition, issues such as interconnect bottlenecks and costs are known to be resolved.
In addition to Samsung Electronics, companies developing BSPDN include TSMC and Intel. Intel named BSPDN "Powervia". Intel also held a Powervia technical briefing last June. These companies aim to apply BSPDN starting from 2nm process.
【Lansheng Technology Information】Vishay Technology recently announced the launch of a new reflective light sensor-VCNT2030 for industrial, computer, consumer and mobile applications. Compared with previous generation solutions, VCNT2030 is more space-saving, while having a higher current transfer ratio (CTR), longer sensing distance and lower power consumption, resulting in improved performance. The product is configured with a vertical-cavity surface-emitting laser (VCSEL) and a silicon photodiode in a tiny 1.85 mm x 1.2 mm x 0.6 mm surface-mount package.
The recently released device adopts a compact structure, the emission source and the detector are in the same plane. The VCSEL offers excellent internal crosstalk rejection due to its narrow ±17° emission angle, which also improves proximity performance behind the cover glass. The analog output signal of the VCNT2030 photodiode depends on the amount of light emitted by the VCSEL and reflected from objects in the sensor's field of view. The sensing distance of the device is 15 mm, which is three times that of the closest competitor on the market.
Due to its small size, the VCNT2030 saves more than 40% PCB space compared to previous generation devices, enabling the sensor to be used in industrial infrastructure, home and building control, laptop and desktop computers, home appliances, consumer electronics and metering applications space-saving optical switching solutions in the field; optical encoding for motor control in e-bikes, golf carts, tractors, and harvesters; and paper detection in printers and scanners. In these applications, the low 8 mA drive current of the device's VCSEL is sufficient to provide the same performance as a solution using a 20 mA infrared emitter, resulting in significantly lower power consumption.
The sensor has an emission wavelength of 940 nm and a typical output current of 2.5 mA, which means a typical CTR of 31% under test conditions. This value is more than 100% higher than the previous generation solution and the closest competitive sensor. The device has a reflow moisture sensitivity level (MSL) of 3 per J-STD-020. It is RoHS compliant, halogen-free, and meets Vishay Green Standards.
【Lansheng Technology Information】ROHM has developed a new dual MOSFET product that integrates two 100V withstand voltage MOSFET*1 for fan motor drive applications in communication base stations and industrial equipment. The new products are divided into two series, "HP8KEx/HT8KEx (Nch+Nch) series" and "HP8MEx (Nch+Pch*2) series", with a total of 5 new models.
Fan motors are usually driven by multiple MOSFETs. In order to save space, there is an increasing demand for dual MOSFETs that integrate two chips. Against this background, ROHM has developed Nch and Pch MOSFET chips using a new process, and has developed a new series of products that achieve the industry's ultra-low on-resistance by adopting a backside cooling package with excellent heat dissipation performance.
The new product achieves the industry's ultra-low on-resistance (Ron)*3 (HSOP8: 19.xn--6m-fcc, HSMT8: 57.xn--0m-fcc for Nch+Nch products) by adopting ROHM's new technology and backside cooling package. Compared with ordinary dual MOSFETs, the on-resistance is reduced by up to 56%, which is very helpful to further reduce the power consumption of application equipment. In addition, by packaging the two chips in one package, the mounting area can be reduced, which helps further save space for application equipment. For example, if a product in HSOP8 package replaces two single MOSFETs (TO-252 package with only one built-in chip), the mounting area can be reduced by 77%.
Currently, ROHM is expanding the withstand voltage lineup of dual MOSFETs for the industrial equipment field, and is also developing low-noise products. In the future, by continuing to help various application products to further reduce power consumption and save space, we will continue to contribute to solving social issues such as environmental protection.
【Lansheng Technology Information】With the rapid development of Southeast Asia, Singapore has gradually become the focus of the semiconductor industry. Important semiconductor manufacturers such as Micron, Intel, and UMC have increased investment in Southeast Asia in recent years to expand their market presence. Even TSMC has reported that it plans to set up its first 12-inch fab in Singapore.
In recent years, Intel has actively deployed in Southeast Asia. It not only has a factory in Singapore, but also held the Intel Solutions Day event in Vietnam in July this year. The event invited partners from the Taiwanese ecosystem to showcase solutions such as the adoption of 4th generation Intel Xeon Scalable processors.
Intel said it regards Vietnam as an important part of its global supply chain layout. Over the years, Intel has aggressively promoted semiconductor packaging and testing in this market. Vietnam not only has a complete information industry supply chain, but also enjoys a demographic dividend, which provides a good environment for the investment and development of the information industry and has great potential for future development.
Micron is a major memory manufacturer. Although it has adjusted its capital expenditures in fiscal years 2023 and 2024, significantly reduced operating expenses in fiscal year 2023, and suspended stock repurchases, its layout in Singapore is still advancing steadily.
It is understood that Micron is currently mass producing 232-layer NAND chips at its fab in Singapore. These chips will first be put on the market in the form of packaged particles to supply Micron's Crucial SSD consumer product series, and more product and supply information will be released in the future.
【Lansheng Technology Information】ON Semiconductor recently released the Elite Power simulation tool and the PLECS model self-service generation tool (SSPMG). The Elite Power simulation tool is paired with PLECS to provide customers with a convenient online environment. Elite Power simulation tools bring several new features and advantages over the current state of the art. SSPMG is a breakthrough tool that increases the flexibility and accuracy of models used in PLECS simulations.
The PLECS model contains loss look-up tables (based on manufacturer data sheets) and thermal chains in the form of Cauer or Foster equivalent networks. During simulation, PLECS uses loss tables to interpolate and/or extrapolate to obtain bias point conduction and switching losses for circuit operation.
The most common method for measuring switching losses is the double pulse tester. ON Semiconductor's advanced dual-pulse testers are valuable for measuring lowest (or device) losses with negligible effect on parasitic elements, allowing comparisons across die sizes, RDS(ON) values and packages. However, data sheet loss values will not reflect the observed losses when customers evaluate losses in their actual applications. After further analysis, it became clear that the datasheet based PLECS model was not representative of the end user's application. Customers had to make a trade-off between cost and performance, so the inductors and capacitors chosen were not as ideal as those found in ON Semiconductor's double-pulse tester. Only if the user's application has the same parasitic environment as the manufacturer's datasheet double-pulse test environment, the simulation can be performed using the standard manufacturer-supplied PLECS model.
A better way to evaluate the customer's actual losses is to introduce the customer's real board and component (such as inductors and capacitors) parasitic elements into the dual pulse tester setup and adapt it to the specific customer application. Since this can be a very difficult measurement task, ON Semiconductor has launched SSPMG for this purpose, enabling customers to design PLECS models according to their application environment in a virtual prototyping environment. While all simulation-based virtual environments are only as good as the underlying model, ON Semiconductor's highly accurate physics-based models are the SSPMG engine customers can rely on to achieve high-fidelity PLECS models.
【Lansheng Technology Information】Renesas Electronics has agreed to acquire French cellular IoT chipmaker Sequans Electronics for US$249 million. The deal is expected to close in the first quarter of 2024, subject to formal approval by the French Works Council, followed by local tax and regulatory approvals. The $249 million valuation includes all stock holdings by all shareholders, including U.S. equities, while also including net debt.
Renesas will integrate Sequans' cellular IoT products and IP into its microcontrollers, microprocessors, analog and mixed-signal front-end products. A statement from the two companies said: "This acquisition will allow Renesas Electronics to immediately expand its business scope into the wide area network (WAN) market covering a wide range of data rates. It will also strengthen Renesas Electronics' already rich personal area (PAN) and local area network (LAN) connectivity product portfolio."
The acquisition is the latest in a spree of acquisitions by Renesas Electronics, which has previously acquired UK power management and industrial IoT specialist Dialog (August 2021), Israeli company Wi-Fi chipset and software provider Celeno (December 2021) and Austrian near-field communication (NFC) chipmaker Panthronics (June 2023). Founded in 2003, Sequans focuses on cellular-based Low Power Wide Area Network (LPWAN) solutions, especially NB-IoT and LTE-M solutions, and IoT hardware based on higher power LTE-4G and 5G. The two parties have been collaborating since 2020 to combine Renesas' embedded processors and analog front-end products with Sequans' wireless chipsets for massive IoT and broadband IoT applications.
【Lansheng Technology Information】Infineon has launched a 62 mm half-bridge and common-emitter module product portfolio with 1200 V TRENCHSTOP™ IGBT7 chips. With a maximum current specification of up to 800 A, the module expands Infineon's product portfolio in a proven 62 mm package design. The improvement of current output capability not only provides maximum flexibility for system designers when designing solutions with higher rated current, but also provides higher power density and better electrical performance. The new modules have been specially developed to meet the needs of centralized solar inverters as well as industrial motor drives and uninterruptible power supplies (UPS). In addition, it is widely applicable to electric vehicle charging stations, energy storage systems (ESS), and other new industrial applications.
Based on the new micro-trench technology, the static losses of the 62mm module series with 1200 V TRENCHSTOP IGBT7 chips are much lower than those of modules with IGBT4 chipsets. These features greatly reduce losses in applications, especially in industrial motor drives operating at moderate switching frequencies. The oscillation behavior and controllability of the IGBT have also been improved. In addition, the new power modules have a maximum overload junction temperature of 175°C.
A robust nickel-plated copper base plate and nut main terminals ensure sufficient mechanical strength for the 62 mm module package. The main terminal is located in the center of the package, making it ideal for parallel circuits and three-level topology configurations due to its low inductance DC link connection. Standard package design and dimensions make this series compatible with previous module versions. In addition, all modules are available with Infineon's proven pre-coated Thermal Interface Material (TIM).