Recently, the chairman of the storage manufacturer Winbond, Jiao Youjun, said that Winbond began to feel the decline in memory sales in the second quarter of 2022. After eight quarters, it gradually felt the increase in sales in the second quarter of this year. It is expected that the quantity will come first, and the price will follow later. The memory industry will enter an upward cycle in the next two years. 2025 will be a significant upward cycle year. The market conditions next year can be described as "significantly optimistic". In terms of terminal applications, Chen Peiming said that the PC terminal market is expected to grow by 5% to 10% this year, and smartphone sales are expected to increase by 5% year-on-year. Netcom benefits from the increase in WiFi 7 applications. The second half of the year can be expected to be optimistic. The real fermentation time will fall next year, and there will be better growth performance. As for automotive and industrial products, the company is optimistic that the automotive market will contribute more this year, and the market conditions in the industrial field will enter the correction at the latest. The current performance is still weak and needs to be adjusted for a period of time.
Beneony
Wholesale Appliances, Electrical, and Electronics
Technology as the core and service as the orientation!!!
About us
We are a global electronic components supplier and exporter specializing in CPU, Memory, Chips, LCD Panels, Active and Passive Components. Not only do we have high-quality supply channels, we also specialize in finding viable supply options for hard-to-find electronic parts. Committed to providing procurement services for various electronic components for domestic and foreign EMS, OEM factories and electronic component dealers. Through efforts and persistence in the electronics industry, Beneony has been recognized and supported by many new and old customers.
- Website
-
www.beneony.com
External link for Beneony
- Industry
- Wholesale Appliances, Electrical, and Electronics
- Company size
- 11-50 employees
- Type
- Privately Held
- Founded
- 2014
Employees at Beneony
Updates
-
According to media reports, according to the latest data, the Chinese market has become the largest export destination for Japanese chip manufacturing equipment for three consecutive quarters, accounting for more than 50%. In the first quarter of this year, Japan's exports of semiconductor manufacturing equipment and related parts and flat panel display manufacturing equipment to China reached 521.2 billion yen, an increase of 82% year-on-year, the highest level since 2007. China has become the largest market for Japanese chip manufacturing equipment! It has accounted for more than 50% for three consecutive quarters According to data from the General Administration of Customs of China, from January to April 2024, the amount of semiconductor manufacturing equipment imported by China from Japan continued to grow, with year-on-year growth rates of 53.06%, 7.45%, 56.82% and 42.19% respectively. China has become the largest market for Japanese chip manufacturing equipment! It has accounted for more than 50% for three consecutive quarters Some analysts believe that the special demand for production line replacement caused by Sino-US trade frictions and the growing demand for automotive electronic components and third-generation semiconductor products from new products such as new energy vehicles are the main reasons. At present, Japanese companies that produce semiconductor manufacturing equipment mainly include Nikon and Canon, which occupy an important position in the global lithography machine market. With the rapid development of China's new energy vehicle industry, the demand for 28nm process chips has increased, further promoting the growth of Japanese semiconductor manufacturing equipment exports to China. China has become the largest market for Japanese chip manufacturing equipment! It has accounted for more than 50% for three consecutive quarters In addition, data from the Semiconductor Industry Association (SEMI) showed that although global semiconductor equipment sales shrank by 2% year-on-year in the first quarter of 2024, sales in the Chinese market increased by 113% year-on-year to US$12.52 billion, maintaining the world's largest chip equipment market position for four consecutive quarters.
-
On March 3, AMD CEO Lisa Su gave a speech at the COMPUTEX Taipei International Computer Show and released the latest AI chip: MI325X. Lisa Su said that the MI300 series has always been AMD's rapidly developing star product, and the new generation of MI325X has inherited this fine tradition. This chip is not only equipped with advanced HBM3E high-bandwidth storage technology, but also adopts the new CDNA3 architecture to ensure its excellent performance. In terms of performance, MI325X is an industry leader. It is equipped with up to 288GB of HBM3E storage and can provide an amazing bandwidth of 6TB per second. Compared with NVIDIA H200, MI325X not only has an advantage in memory capacity and bandwidth, which is almost twice as high, but also has a computing speed that is 30% faster. What's more gratifying is that this chip is also very competitive in terms of cost performance and is expected to be officially supplied in the fourth quarter of this year. In addition to MI325X, AMD also looks to the future. It is revealed that the company plans to launch a new generation of MI350 series chips in 2025. This chip will use cutting-edge 3nm process technology and be based on a new architecture design. The MI350 series will integrate 288GB of HBM3E memory and support FP4/FP6 data formats, making it 35 times faster than the existing MI300 series chips in inference operations. Previously, there were rumors that AMD would adopt Samsung's 3nm process technology. Su Zifeng made it clear at COMPUTEX that TSMC is a solid partner of AMD, and the two sides currently have multiple 3nm process product cooperation projects.
-
NVIDIA rose to US$954, once again setting a record high, and its market value has reached US$2.35 trillion. According to the Financial Associated Press, a comparison chart produced by industry insiders shows that NVIDIA's current market value is equivalent to the sum of "TSMC + ASML + AMD + Qualcomm + Applied Materials + Texas Instruments + Micron + Intel". NVIDIA will announce its performance report for the first quarter of fiscal year 2025 in the early morning of Thursday morning Beijing time. Wall Street expects Nvidia's revenue to increase by 240% to US$24.6 billion in the first fiscal quarter, and net profit to increase by 540% to US$13.1 billion. After this week, can Nvidia further increase the number of semiconductor companies on the right? We'll see.
-
When selecting TVS diodes and ESD electrostatic diodes, you should make a judgment based on the actual needs of the circuit and the application environment: ●If the circuit needs to face overvoltage and electrostatic discharge threats from both power and signal lines, then TVS diode is a better choice because it can provide bidirectional protection and respond quickly to transient voltages. ●If the circuit mainly faces the threat of electrostatic discharge and the protection requirements for transient voltage are not particularly high, then an ESD electrostatic diode may be more suitable because it is specifically designed for electrostatic discharge and has less impact on the normal operation of the circuit. In practical applications, in order to improve the reliability of protection, TVS diodes and ESD electrostatic diodes are usually used together to achieve multiple protection of the circuit. In addition, other protective components should also be considered, such as gas discharge tubes (GDT), resettable fuses (self-resetting fuses), EP4CGX75CF23C8N varistors, etc., which can also provide different types of protection against different electrical threats. In short, when considering transient overvoltage protection, TVS diodes and ESD electrostatic diodes are both important protection components. Reasonable selection and configuration of these protection components can significantly improve the system stability and reliability of electronic products.
-
According to the official WeChat account of China Telecom, the Institute of Quantum Information and Quantum Technology Innovation of the Chinese Academy of Sciences (hereinafter referred to as the "Quantum Innovation Institute") successfully delivered a 504-bit superconducting quantum computing chip called "Xiaohong" to Guodun Quantum. , this move aims to verify the performance of the kilobit measurement and control system independently developed by Guodun Quantum. The advent of this chip has undoubtedly set a new record for the number of superconducting qubits in China. In the future, it is planned to be open to the world through the "Tianyan" quantum computing cloud platform of China Telecom Quantum Group to promote the international exchange and development of quantum computing technology. cooperate. Measurement and control systems and quantum computing chips are indispensable core hardware components of quantum computers. Among them, the measurement and control system ensures the accurate generation, transmission and processing of signals through precise interaction with the quantum computing chip, thus profoundly affecting the overall performance of the quantum computer. In order to comprehensively verify various performance indicators of the large-scale measurement and control system, the Quantum Innovation Institute specially customized and developed this 504-bit "SnapHong" quantum computing chip. Thanks to the Quantum Innovation Institute's deep R&D and processing capabilities in the field of superconducting quantum computing chips, this chip not only integrates more than 500 bits, but also has qubit lifetime, gate fidelity, gate depth, read fidelity, etc. In terms of key indicators, it is expected to reach the chip performance of internationally renowned quantum computing cloud platforms such as IBM, fully meeting the verification needs of kilobit measurement and control systems.
-
According to media reports, Intel disclosed in its Gaudi 3 AI chip white paper that it is preparing to launch a "special edition" Gaudi 3 to the Chinese market. Gaudi 3 specially supplied to China includes an OAM-compatible mezzanine card named HL-328 and a PCle accelerator card named HL-388. HL-328 will be launched on June 24, and HL-388 will be launched on June 24. Launched on September 24th. Compared with the original version, the China special edition Gaudi 3 has the same 96MB SRAM on-chip memory, 128GB HBM2e high-bandwidth memory, a bandwidth of 3.7TB/s, and a PCIe 5.0 x16 interface and decoding standard. However, due to the export control of AI chips in the United States, its comprehensive computing performance (TPP) needs to be lower than 4800 before it can be exported to China. This also means that the 16bit performance of the China special version of Gaudi 3 cannot exceed 150 TFLOPS. The performance of the original Gaudi 3 on FP16/BF16 can reach 1835 TFLOPS, so the Chinese special version of Gaudi 3 may eventually need to reduce its AI performance by about 92% to comply with US export control requirements. However, the reduction in performance has also significantly reduced its power consumption. According to exposed data, the TDP of the PCIe card and OAM card of the China special version of Gaudi 3 are both 450 watts, while the performance of the original version is 600 watts and 900 watts respectively.
-
According to media reports, Musk’s latest prediction in an interview is that artificial intelligence may surpass the smartest humans next year or before 2026. “If you define general artificial intelligence as having intelligence that exceeds that of the smartest human being, I think it could be next year, or two years at the most,” Musk said. Musk believes that the chip supply shortage faced by the AI industry in the past is being alleviated, but new challenges have emerged, and the demand of AI models for other equipment in data centers and power grids is still increasing. Last year, the main challenge for the AI industry was the shortage of chips, and it was difficult for people to obtain enough Nvidia chips; this year, the problem gradually turned to the supply of voltage transformers. "It is expected that within one to two years, the main bottleneck in the development of the AI industry will be power supply." Musk said. Musk also revealed that xAI is training a second-generation Grok model. "We think it may be better than GPT-4," and it is expected to complete this stage of work in May.