TSMC begins trial production of 2nm chips for Apple and Nvidia

According to foreign media reports, the global foundry leader TSMC has not only started to develop a 2-nanometer process, widening the gap with competitors, but also recently began to prepare for trial production of 2-nanometer products for Apple and NVIDIA. In addition, in order to further develop the 2nm process technology, TSMC will send about 1,000 R&D personnel to work in the Fab 20 fab, which is currently under construction.

According to the Patently apple report, Samsung Electronics will use GAA technology to start mass production of 3nm process chips in June 2022, six months ahead of TSMC, becoming the world’s first company to mass produce this process technology. Under the impact of Samsung’s preemptive strike, TSMC executives have repeatedly disclosed the development plan of 2nm process technology, forming an advanced process development competition. In addition, it is reported that the current TSMC is still developing the technology of power supply on the back of the chip, and the goal is to use this technology by 2026.

In addition to TSMC, Intel, the major processor manufacturer that previously announced plans to re-enter the foundry business in 2021, has also joined the advanced process R&D race. The US semiconductor giant announced the technical development, test data and roadmap of its chip backside power supply solution PowerVia in an online event on June 1, local time, and began to expand its influence in the foundry industry.

Additionally, Intel has set a goal of advancing its foundry technology to a 1.8nm node in the second half of 2024. In March, the company laid out a plan to reach volume production on its 1.8nm process technology through a partnership with ARM. However, market participants also have some uncertain views. They believe that even if Intel succeeds according to the roadmap, it will still be a big challenge for the company to eventually achieve balance of payments.

The report also explained the situation of Samsung, another competitor of TSMC. Kyung Kye-hyun, president of Samsung DS Division, said in a speech in early May that Samsung plans to surpass TSMC. The goal is to use the 2nm process of GAA technology earlier than TSMC. start.

In fact, in addition to advanced manufacturing processes, TSMC also maintains its technological leadership through advanced packaging technologies. Not long ago, TSMC announced the official opening of the sixth advanced packaging and testing factory, becoming the first TSMC to achieve the integration of front-end and back-end process 3D Fabric integration automation advanced packaging and testing factory and testing service factory. At the same time, it is also preparing for the mass production of TSMC-SoIC (system integrated chip) process technology. The opening of the sixth advanced packaging and testing factory will enable TSMC to have a more complete and flexible production capacity planning for SoIC, InFO, CoWoS, advanced testing and other TSMC 3DFabric advanced packaging and silicon stacking technologies, and also bring higher production. Yield and performance synergies.

looking for IGBT modules such IRG7T200CL12B and IRG7T150CL12B, igbtexpress.com have more than 300 thousand parts you can choose.

Samsung Electronics will develop 4F2 DRAM storage unit

With the slowdown of Moore’s law, the DRAM process has also entered the technical bottleneck period, and the expansion speed of DRAM has slowed significantly. Therefore, 3D DRAM has become a new path for storage manufacturers to break through the higher limit of DRAM technology. Recently, according to Korean media The ELEC, Samsung Electronics has formed a development team in its semiconductor research center to mass produce 4F2 DRAM. The goal of the development team is to apply the 4F2 DRAM storage unit to the DRAM process below 10 nanometers, and at the same time, it solves the problem of current technology facing the limit of scratching lines.

The reason for the slowdown in DRAM process breakthroughs is the simple structure of the storage unit -composed of a capacitor for storing charge and a transistor for access capacitors. To solve the contradiction between the slowdown of DRAM’s expansion and the demand for large capacity, the main solution of the industry is currently subverting this structure, while adding some special materials to promote DRAM process innovation.

Earlier, the industry mainly increased the density of DRAM chips by reducing the width of the route. The smaller the width of the line, the more the transistor, the higher the integration, the lower the power consumption, and the faster the speed. Although this method does have a certain effect, after the line width enters 10nm, the problem of physical restrictions such as power leakage and interference of capacitors follows.

Of course, the industry has also introduced new materials and new equipment such as HIGH-K materials and extremely ultraviolet (EUV) equipment to solve this problem. However, in the manufacture of 10nm or more advanced small chips, some of the existing technologies can no longer overcome DRAM physical limitations. With the increasing contradiction between DRAM’s technical technology and insufficient supply of DRAM, 2D DRAM risen to 3D DRAM has gradually become a consensus for the industry’s pursuit of technological breakthroughs.

The 3D DRAM is a new type of storage method that stacked the storage unit (Cell) to the top of the logical unit, so as to achieve higher capacity in the unit wafer area. In principle, 3D DRAM can effectively solve the current dilemma of plane DRAM. At the same time, in terms of cost, the 3D stack technology used by 3D DRAM will achieve reusable use of storage capacitors, which can effectively reduce unit costs. It can be seen that DRAM’s development from traditional 2D to 3D three -dimensional will be the future development trend.

Recently, according to the foreign media “Businesskorea”, Samsung’s main semiconductor leader recently stated at the semiconductor meeting that is accelerating the commercialization of 3D DRAM, and 3D DRAM is a way to overcome DRAM’s physical limitations. It will change the game of the memory industry rule. At the same time, 3D DRAM is considered the future growth momentum of the semiconductor industry.

Since 2022, the depression of the electronic consumer market has allowed the memory market to enter the “cold winter”, but the demand for memory in other fields such as automotive electronics and AI servers is still strong, especially the demand for high -performance storage such as HBM brought by ChatGPT will accelerate the acceleration. DRAM 3D development. TrendForce Ji State Consultation predicts that AI demand continues to drive the growth of HBM memory, and it is estimated that the compound annual growth rate of HBM markets from 2023 to 2025 is expected to grow to more than 40 ~ 45%.

The ELEC reports that if the structure of the Samsung 4F2 DRAM storage unit is successful, compared with the existing 6F2DRAM storage unit structure without changing the node, the chip DIE area can be reduced by about 30%. limit. The 4F2 structure is a unit structure technology that the DRAM industry failed to commercialize about 10 years ago. It is said that the process is quite difficult. Data show that 6F2 can reduce the area by 25-30%compared with 8F2.

It is reported that at present, the industry has the design of 8F2 and 6F2 DRAM units, including the unit including 1T (crystal tube) and 1C (capacitor). This 1T+1C unit design will be used for the DRAM unit design of several generations of DRAM in the future. However, due to the restrictions of process and layout, DRAM manufacturers have been developing the 4F2 unit structure, such as 1T DRAM or a capacitor DRAM prototype, as one of the next candidates for extended DRAM technology.

At the 2021 IEDM, the team of Researcher Li Ling Institute of Microelectronics Institute of the Chinese Academy of Sciences and the Huawei/Hisilicon team first proposed a new CAA. This structure effectively reduces the area of the device and supports multi-layer stacks. By directly connecting the upper and lower CAA devices, the size of each storage unit can be reduced to 4F2, which makes Igzo-DRAM have a density advantage.

In January 2023, Academician Liu Ming team of the Microelectronics Key Laboratory of the Institute of Microelectronics of the Chinese Academy of Sciences studied the impact of the second -layer device stacking the front layer process of the front layer of the front layer on the basis of the vertical ring channel structure (CAA) Igzo FET. The reliability of CAA IGZO FET in the 2T0C DARM application.

looking for IGBT modules such IRG7T200HF12B and IRGTI090U06, igbtexpress.com have more than 300 thousand parts you can choose.

Sino-US trade has a huge impact on 50% of the world’s semiconductors

The first wave of trade wars did not affect the world information industry chain, but after the second wave, tariffs were added to the 267 billion US dollars of products, including the communication products, and the global communications industry will be affected. If the event evolves to this point, it will affect the global supply chain. If the cost becomes higher, the price will increase and the demand will decrease. The Sino-US trade war will have a huge impact on 50% of semiconductor traders, and it will also hurt the international economy.

1 Trade war continues to plague semiconductor supply chains

The Trump administration’s tens of billions of dollars in tariffs on China’s imports and Beijing’s retaliatory tariffs have raised concerns about trade wars that will have an impact on China’s chip sales and create chains through supply chains. reaction.

If the scale of Sino-US trade wars is upgraded again, it will threaten to spread more Chinese system manufacturers.

Although it will affect the overall output of China’s electronics manufacturing industry, it will benefit to have the strength to complete domestically replaced electronic component manufacturers.

In the context of the tight global electronic component supply chain, compared with the world’s advanced level, the Chinese chip industry still has a huge gap.

When supply is tight, foreign companies with monopoly positions can take the opportunity to raise prices, while most domestic manufacturers do not have bargaining power.

The current Sino-US trade war continues to escalate and has been recognized as a protracted war for trade.

A major driver of the US trade war is to curb the development of China’s manufacturing industry and crack down on the “Made in China 2025″ plan. The semiconductor industry has not escaped the trade war.

This risk is now exposed and not entirely bad. The Sino-US trade war will force more system manufacturers to cooperate with domestic component manufacturers and spend more energy to tap the domestic market demand. This is an opportunity in risk.

2 The long-term goal of the US is still to promote local supply chain investment

The long-term policy goal of the United States will still be to promote the industry to invest and set up factories in the United States.

The United States originally intended to increase the cost of US companies’ overseas production through punitive tariffs through means of trade wars, undermining the comparative advantage of the original advantages of mainland China and other countries, enabling US companies to return to the United States to invest and expand their production capacity to meet their domestic markets. demand.

In this trade war, the biggest threat to mainland China comes from the semiconductor supply and software licenses of the US and other industries.

In this trade conflict, taking the ZTE incident as an example, the biggest bargaining chip in the United States is to implement the prohibition of component output and software licensing for Chinese mainland players. This will cause an immediate impact on mainland China, which is equivalent to slamming mainland China. The throat of the operator.

From the long-term strategy of the United States and China, the United States wants to strengthen downstream manufacturing capacity, combine existing software and semiconductor advantages, and build a complete domestic supply system, thereby leaving the company’s investment in the country and creating more employment opportunities.

Mainland China is eager to get rid of dependence on imported semiconductors and actively establish a supply chain for the local semiconductor industry.

3 China needs time to test for monopoly

However, in this trade conflict, not only the United States wants to establish more industrial supply chains in the country, but China will also actively build local supply chains after this trade conflict.

Unlike the United States, which is dominated by downstream manufacturing industries, mainland China is more active in upstream electronic components such as semiconductors and other industries and software industries.

4 Domestic integrated circuit market in recent years

Can be found to have the following characteristics

1 Although the market is large and self-sufficient, from the data of China’s integrated circuit import and export in 2013 to the year, in addition to the small amount of domestic chip occupancy in the communication equipment, the other fields are basically zero;

2 Although there are many inputs and slow development, from 2010 to 2017, the state’s investment in fixed assets of integrated circuits has reached 78 billion from 488.8 billion, and the corresponding chip industry has also shown good development momentum, but it is still difficult to meet the growing domestic demand. According to SEMI data, in 2017, Chinese companies can only meet 26% of local chip demand;

3 Insufficient revenue and low profit. Due to the high risk of high-tech industries, ZTE is used as an example. Its annual report shows that the revenue only accounts for about 20% of the company’s profit, which leads to the fact that although the state takes the lead in capital investment, there is no profit-driven power. Social capital is still reluctant to enter.

looking for obsolete electronic components such MK20DX256ZVLK10 and 2LHT-L5, lincomponents.com have more than 300 thousand parts you can choose.

IDC predicts that there will be nearly 1.3 billion global smart home devices in 2022

According to IDC’s global quarterly smart home device tracker, the global smart home equipment market is expected to grow 31% year-on-year in 2018, reaching 644.9 million. The entire smart home market, including smart speakers, video entertainment products, connected lighting, smart thermostats and home monitoring/security products, is expected to have nearly 1.3 billion devices by 2022, with a compound annual growth rate of 20.8% over a five-year period. The fastest-growing category, with a compound annual growth rate of 39.1% in five years, nearly 100 million units in 2018, and 235 million units by 2022, will become smart speakers such as Amazon Echo and Google Home.

Jitesh Ubrani, senior research analyst at IDC Mobile, said: “Although dedicated smart speakers with built-in voice assistants will be popular throughout the forecast, we have seen significant changes in this market.” “Many new devices inside and outside the home are now Includes built-in voice assistants. These devices not only help increase the number of touchpoints available to end users, but also help each assistant grow and gain insight into the content, time and manner of the user’s various tasks by expanding coverage.”

The privacy and security of smart home devices remains one of the key factors, as IDC’s recent consumer Internet of Things. However, IDC remains optimistic about the market’s prospects, as the convenience of smart home devices often exceeds concerns.

Adam Wright, senior research analyst at IDC Consumer IoT Program, said: “When considering the deployment of smart home devices, most consumers put security and privacy first.” “In addition to having enough equipment to bring the device to market In addition to security solutions, vendors must respond to consumer concerns by appropriately educating them about privacy protections and increasing their awareness of the benefits of shared data, which can ultimately provide better data over time. Equipment and services.”

Smart home equipment category highlights

Video entertainment products are mainly composed of smart TVs and digital media adapters, such as Google’s Chromecast, Apple TV, Fire TV and Roku devices. As the center of the living room, this category is expected to grow. It is expected to reach 457.5 million units by 2022, with a five-year compound annual growth rate of 10.9%. For many brands and consumers, this category is expected to be the gateway to the smart home ecosystem, so IDC expects significant competition in terms of content and price in this category.

Home monitoring/security equipment, from door/window sensors to door locks and IP cameras, is expected to account for 19.4% of the smart home market by the end of 2022. Although this category faces great potential with a five-year growth rate of 27.3%, it is still targeted at a limited consumer base, especially homeowners, as most products still do not provide much value to consumers living in apartments.

Although smart speakers have become the focus of attention today, it will become the third largest category in 2022, as many vendors offer products that support Amazon Alexa or Google Assistant. Siri, Cortana and Bixby are also expected to build a wide range of usability, although none of these assistants will challenge Google or Amazon’s position in terms of unit shipments. Although a large part of the market will not be able to run on these platforms, IDC expects that most of these smart speaker markets will flow to China, and established companies such as Xiaomi and Alibaba have already taken a large share.

Connecting lighting, thermostats and other products will complement the smart home market, providing users with a complete ecosystem of products and services to control and monitor their homes. The five-year compound annual growth rate for these categories is expected to reach 26.9%, while lighting products (such as Philips Hue) account for almost one-third of shipments. Although thermostats are popular in North America, they will face more growth in other parts of the world, as many countries do not use central HVAC systems and therefore do not require dedicated thermostats.

looking for obsolete electronic components such MV2201 and XC17V02PC20I, oemsemi.com have more than 300 thousand parts you can choose.

Micro-LED surface brightness detection based on microscopic imaging system

The research team of Xiamen University and Taiwan’s Hsinchu Jiaotong University (Guo Haozhong, Wu Tingzhu, Zhao Qingwei) recently developed a micro-LED surface brightness detection system based on microscopic imaging system, which can quickly measure the position of micro-LED array at any position during operation. Absolute brightness value.

LEDs have existed for more than half a century as an active self-illuminating device. Its low power consumption, low operating voltage, high illumination, long operating life and stable performance make it a very popular application. With the development of LED technology and the growing demand for micro-integration in various industries, micro-LED array devices have emerged. By further reducing the size of the LED chip, hundreds or thousands of micro-LEDs can be fabricated in the millimeter range to form a micro-LED array. micro-LEDs have potential applications in many areas, including micro-LED displays, high-speed parallel visible light communications, and high-voltage lighting chips.

In the display field, compared with LCD and OLED display technologies, micro-LED displays have the advantages of high luminous efficiency, high brightness, high contrast, and short response time. However, due to heat dissipation, aging and other problems, the imaging defects of the display affect the imaging quality, so the defect detection of the display is necessary. However, current measurement measurements remain at a macro level and cannot be used to accurately measure micro LED arrays, and the method used to evaluate the brightness of micro-LED arrays focuses on the measurement of total brightness, failing to quickly detect bad pixels in the micro LED array. . Therefore, it is very necessary to find a fast and accurate method to check the surface brightness of the micro-LED.

looking for obsolete electronic components such FP40R12KT3G and BCR30GM, igbtexpress.com have more than 300 thousand parts you can choose.

Japan Fujitsu released a new super-computer “brain”

Fujitsu Corporation of Japan officially released the “A64FX” processor, the core component of Japan’s next-generation supercomputer.

The “A64FX” processor features ultra-low power consumption and high reliability with at least 2.7 trillion floating point operations per second. Typically, supercomputers use tens of thousands of processors. Fujitsu said in its press release on its website that the next-generation supercomputer with the “A64FX” processor will perform up to 100 times faster than the current Japanese supercomputer “Kyoto.”

The next-generation supercomputer of “Kyo” is jointly developed by Fujitsu and the Japan Institute of Physical Chemistry, and the research and development cost is about 110 billion yen (about 6.8 billion yuan). This project has received financial support from the Japanese government. Fujitsu plans to run this new supercomputer by around 2021.

According to Japanese media reports, the Japanese supercomputer “Beijing” won the global super-calcin 500 list in 2011, but has fallen to the 16th in the world in June this year, Japan is looking forward to regaining the global super by developing the next generation of supercomputers. Count the top.

In the list of the world’s top 500 supercomputers released in June this year, the “superior” of the US supercomputer surpassed China’s “Shenwei·Taihu Light” ranked first, which is the first time that the US supercomputer has returned to the top.

looking for igbt modules electronic components such BSM300GB120DLC and BSM50GX120DN2, igbtexpress.com have more than 3 thousand parts you can choose.

BYD Electronic Control Latest Technology Progress: Will Increase Research on IGBT Modules

Analysts of the New Sijie Industry Research Center believe that in the future, BYD will increase the research and development of IGBT modules, on the one hand to improve the electronic control performance, to meet the needs of new models, and on the other hand to expand the market size of IGBT modules to achieve external sales.

BYD Co., Ltd. has always attached importance to the development and innovation of new products and new technologies, and regards the development of new products as an important guarantee for the company to maintain its core competitiveness. BYD has always attached great importance to the research and development of new products and new technologies. It has a strong R&D team, built a multi-disciplinary research system and a strong patent reserve, enabling the company to have global leading technology in the fields of battery, motor and electronic control. strength. The daily work of the electronic control system is to distribute the energy to the four wheels through electronic signals, and at the same time, the sensors throughout the vehicle will feedback and exchange the information of each part in time.

In the “2017-2021 China Electric Vehicle Market Analysis and Development Prospects Research Report” released by Xinsijie, the latest technology of BYD Electronic Control is discussed in detail:

In terms of electronic control design, BYD provides the BM3451/3452 series IC independently developed by BYD, which has the characteristics of high precision and high integration. Through monitoring the voltage, charge and discharge current and ring temperature of each battery to achieve battery overcharge, over discharge, over current, temperature protection and other protection functions, external capacitors can be used to adjust overcharge, over discharge, overcurrent protection delay Time. In addition, the BM3451/3452 series is available in a variety of packages to provide customers with customized chip sizes.

BYD has the core technology of VCU/MCU, which is developed by BYD’s internal automotive engineering research institute. The pure electric vehicle controller is VCU and the hybrid power is HCU. In both aspects, BYD has mature technology. The technology collects driver driving signals through sensors, obtains motor and battery information through CAN bus, and shares and calculates. Motor and battery management commands are given via the CAN bus for vehicle drive control, energy optimization control and brake feedback control.

The motor controller is mainly composed of a control circuit, a drive circuit, a housing structure, and various sensors. The most important of these is the IGBT module in the drive circuit, which accounts for about 40% of the cost in the entire electronic control system. BYD is a vehicle manufacturing enterprise and produces its own electronic control products. Due to the more complicated application conditions of IGBT chips for new energy vehicles, high temperature, high humidity and high vibration, as well as assembly volume and cost, are very strict, which gives BYD’s IGBT. The development and production of chips put forward higher requirements.

Analysts of the New Sijie Industry Research Center believe that in the future, BYD will increase the research and development of IGBT modules, on the one hand to improve the electronic control performance, to meet the needs of new models, and on the other hand to expand the market size of IGBT modules to achieve external sales.

CAST and Achronix use lossless compression IP to support data processing from the data center to the edge

Integrate CAST’s compressed intellectual property (IP) with Achronix’s eFPGA technology in a high-performance, low-power solution to support the movement and storage of big data

Achronix Semiconductor Corporation, a leading provider of field-programmable gate array (FPGA)-based hardware accelerators and embedded FPGAs (eFPGA), today announced that it is focused on providing semi-conductor IP knowledge for semiconductor system designers. The property company CAST Incorporated has entered into a partnership; CAST’s high performance lossless compression IP has been implanted to support Achronix’s FPGA portfolio for efficient processing of data transfer between the data center and the mobile edge. SAA7356HL

CAST provides standard hardware implementations for lossless compression tools such as Deflate, GZIP, and ZLIB that are compatible with software implementations for compression or decompression. The hardware implementation provided by the ZipAccel core provides high throughput up to 100Gbps with very high compression performance and low latency. Coupling it with Achronix’s Speedcore eFPGA technology enables a high-performance, low-power solution that can more easily move and store large data.

With the explosion of applications with analytics, more and more information is being transmitted through bandwidth-limited communication channels that are widely used in many scenarios from automotive systems to large financial institutions. The cost and power consumption of transmitting data is becoming more and more important. The compression function implemented using Achronix eFPGA can minimize power consumption and maximize network capacity. In a custom SoC, the combination of CAST Compression IP and Speedcore eFPGA IP can effectively increase the achievable throughput; in addition, developers can use eFPGA to quickly and efficiently implement data processing algorithms. FTLX8571D3BCL

In order to meet the system’s specific throughput, storage, and latency requirements, the ability to optimize compression algorithms in eFPGA will make this solution an alternative technology for hundreds of application scenarios. This not only increases throughput, but also allows significant savings in expensive memory storage space.

“We are very excited to work with CAST to further enrich Achronix’s partner program,” said Mike Fitton, senior director of product planning and business development at Achronix. “The ability to instantiate CAST’s high-throughput compression core in our eFPGA will support ASICs and SoCs with Speedcore to effectively meet the data services market. Thanks to eFPGA IP as a repeatable workload for specific workloads. Programmed hardware accelerators, which enable new algorithms including compression and data parsing, to be quickly implemented. The high performance of the Speedcore eFPGA coupled with its large market traction make it an ideal choice for this type of application.”

“CAST is pleased to license the core to Achronix’s customers who will benefit from Achronix’s unique architecture, which will provide new algorithms with a high degree of flexibility and timeless capabilities, and enable rapid time-to-market,” CEO of CAST Ltd. Nikos D. Zervas commented. “By saving time for our customers who require these features in Achronix FPGAs and eFPGAs by providing proven IP solutions for the Achronix FPGA toolchain and architecture, the IP has been further optimized to take full advantage of Achronix’s FPGAs. Architecture to accelerate and reduce chip area.”

What is the odds of China winning the semiconductor-free war?

April was an eventful event for China’s electronics industry. First, on April 3, U.S. trade signed an additional 25% import duty on more than 120 items of technology electronic products in China. Then, on April 16, the U.S. Department of Commerce announced that it prohibited US companies from selling to ZTE. Semiconductor products, aging for 7 years.

It is not difficult to see that the United States is using the banner of “unfair trade” to take a further trade war against China. It is hard to imagine how these bans will be fully implemented and how ZTE should deal with them to survive the crisis.

A series of measures in the United States undoubtedly made a splash to China. Due to the rise of artificial intelligence and the Internet of Things, semiconductor chips have ushered in an unprecedented golden era. In 2017 alone, the amount of imported semiconductor chips in China reached 260.14 billion U.S. dollars, and China’s self-determination of semiconductor chips has become a priority. The ZTE event once again highlighted the “dead spots” of China’s semiconductor chips.

To sum up the current situation of China’s semiconductors, perhaps some inspiration. Current status of China’s semiconductors: 1. At present, China is the world’s largest semiconductor consumer country, and consumes 45% of the global market share of semiconductor chips each year. 2. China’s 90% of semiconductor chip consumption depends on imports, and imports far exceed oil prices. 3. Chinese semiconductor chips started relatively late and fell behind the world average. 4. The world’s semiconductor giants have formed a complete ecological chain, and the commercialization of Chinese semiconductor companies is difficult.

A lot of people in the ZTE incident analyzed the result that both China and the United States will suffer. But in reality, the United States is “injured”, but ZTE means “dead.” Since ZTE collapses, the United States can continue to support other foreign communications companies to achieve alternatives and quickly absorb ZTE’s market share.

Under the intervention of both governments, the two parties are likely to reach an agreement. However, if this negotiation lasts for more than two months, ZTE’s inventory will be exhausted and it will have irreversible consequences.

After this incident, Chinese companies were able to learn from their pain and make more independent research and development of chips. However, in order to achieve complete self-supply of semiconductor chips in the short term, it is absolutely impossible for them to focus on the future and have an odds.

China is so powerful. Why is it that a small semiconductor chip is stuck? The fact is, in the semiconductor field, even if you have money, you cannot do what you want.

Originally, China could rely on the acquisition of semiconductor companies in other countries to obtain core interests, but a “Waistler Arrangement” has allowed Chinese semiconductor acquisitions to be rejected again and again. The Wassenaar Arrangement, also known as the Wassenaar Arrangement Mechanism, was established in 1996 after the dismantling of the Paris Coordinating Committee by the world’s major manufacturers of industrial equipment and weapons to control the international nature of conventional weapons and high-tech trade. organization. In this organization, the object of the target blockade mainly includes the major categories of vehicles: traditional weapons such as tank artillery, and weapons of the scientific and technological community, both military and civilian goods and technology. Among the countries that are restricted, there is China.

Back to the semiconductor chip industry, it is because of the Wassenaar Arrangement that it is difficult for China to acquire chip companies from abroad or import many devices in the chip field. While China wants to create its own chip technology, it needs a lithography machine, but China can’t buy any advanced lithography machine equipment and independently research and develop the lithography machine’s R&D experience. The gap is too big. Therefore, it is particularly difficult for China’s semiconductor industry to start.

Even if the “Wassenaar Arrangement” is bypassed, or the photolithography process can be used to achieve the world standard, China is still not complete in the semiconductor chip ecosystem, and the overall progress is slow. In recent years, China has achieved a lot in the semiconductor industry. However, we must admit that there is still a wide gap between China and the world’s advanced semiconductors.

In fact, China’s semiconductor industry will not take short cuts to win, but long-distance travel is long. To achieve the world’s top level, we must change from a fundamental point of view. The most important issue is talent. At present, China’s most in short supply is semiconductor talent to attract and cultivate semiconductor talent; followed by investment and emphasis on technological breakthroughs, and then China should increase semiconductors that encourage long-term investment in patient capital and cyclical investment in patient capital. The opportunity for companies to give trial and error cannot be achieved overnight.

The semiconductor industry is a technology-based field. Therefore, we only have constant innovation and master core technologies before we can have a certain amount of discourse in the international community. From a certain perspective, the ZTE incident is now a good opportunity for China, which can cause reflections on Chinese electronics companies and also promote Chinese industry innovation. BSM50GX120DN2

Intel accumulated more than 50 years of technical data from several scientists in materials physics before releasing the original PC chip; Samsung’s flash memory technology, 3500 engineers worked successfully for 20 years to design and complete the manufacturing process. FS450R12KE3

Therefore, China also needs to maintain long-term patience and hope that they can make major breakthroughs overnight. In the semiconductor industry, such shortcuts will not work. Only by exerting national strength and maintaining unwavering semiconductor R&D determination can we complete the grand industry of semiconductor chips and let us work together for China!

TSMC expects Q2 revenue to be lower than market expectations

Due to the weak smartphone market and rising uncertainty in the cryptocurrency mining market, coupled with the appreciation of the exchange rate of the Taiwan dollar against the US dollar, TSMC estimated that the consolidated revenue in NT in the second quarter will be between 2,277.6 and 230.68 billion yuan. During the period, it fell by 7.0 to 8.2% from the previous quarter, which was lower than market expectations. TSMC also reduced its annual revenue growth target from 10-15% estimated at the beginning of the year to 10%.

Affected by the negative impact of TSMC’s downward revision this year, US stocks earlier in the day, TSMC ADR fell by more than 6%, fell to 39.38 US dollars.

Foundry leader TSMC held a legal briefing yesterday (19th). The company’s consolidated revenue in the first quarter reached US$8.459 billion, a decrease of 8.2% from the fourth quarter of last year and a 12.7% increase from the same period of last year. Affected by the appreciation of the exchange rate against the United States dollar, the NT$ consolidated revenue for the first quarter of the year fell 10.6% to 248.079 billion yuan, still 6.1% higher than the same period last year. SP506CF

TSMC’s co-chief executive Wei Zhejia said that the main reason for the outlook adjustment is that the demand for smart phones is lower than expected, and the demand for cryptocurrency mining ASIC market is also full of uncertainties.

TSMC’s average gross profit margin in the first quarter was 50.3%, and its operating profit rate was 39.0%. It was in line with its previous performance outlook. The single-season parent company’s net profit after tax fell 9.6% to RMB 89.788 billion, a growth of 24.6% compared to the same period last year, and net profit per share. 3.46 yuan, in line with market expectations. TSMC said that 28nm and more advanced process revenue has reached 61% of the total wafer sales.

However, TSMC’s outlook for the second quarter was conservative, mainly due to the weak demand from the smart phone market. It is expected that the consolidated revenue will be between US$7.8 billion and US$7.9 billion, representing a decline of 6.6 to 7.8% from the previous quarter and 10.5% from the same period of last year. ~11.9%. Under the assumption that the exchange rate of the New Taiwan dollar against the US dollar is 29.2 yuan, the consolidated revenue in NT will be between 2,277.6 and 230.68 billion yuan, down by 7.0 to 8.2% from the previous quarter, which is lower than market expectations, but the same period of last year. Compared to still grow 6.5 to 7.9%. BTW58-1500R

TSMC estimates that the gross profit margin for the second quarter is between 47 and 49%, and the operating profit rate is between 35 and 37%. This compares with the first quarter and the same period of last year. TSMC said that the main reasons include the appreciation of the Taiwan dollar and poor product mix.

Regarding cryptocurrency computing market ASIC market changes, TSMC said that related performance in the second quarter is still better than the first quarter, and the second half will be better than the first half, but after the recent cryptocurrency price drop, 28nm ASIC demand turns Weak, strong demand for 20nm and more advanced process ASICs, so this market is still conservative and cautious.

TSMC also revised its outlook for the semiconductor and foundry market this year. TSMC estimates that the annual growth rate of the semiconductor market that does not include memory this year is about 5%, which is only 5-7 percent of the bottom margin estimated at the beginning of the year; the foundry market grew by 8% from last year, slightly lower than the forecast at the beginning of the year. Increase by 9 to 10% annually. TSMC’s revenue in US dollars this year is about 10% higher than last year, lower than the annual increase of 10% to 15% expected at the beginning of the year.

Free Web Hosting