Hot News
The Huawei Kirin 9020 chip makes its public debut, with technology fully under independent control.
Recently, Huawei’s new generation of foldable flagship Mate XTs was launched, and Huawei’s Executive Director and Chairman of the Terminal BG, Yu Chengdong, announced at the launch event that the device is equipped with the Kirin 9020 chip. Yu Chengdong stated that the Kirin 9020 achieves a breakthrough in performance through software-hardware cloud collaboration and system-level deep optimization, resulting in a 36% overall performance improvement, delivering extraordinary performance.
The chip’s CPU consists of 1 large core at 2.5GHz, 3 medium cores at 2.15GHz, and 4 small cores at 1.6GHz, with the GPU being Maleoon 920. The Kirin 9020 achieves full self-research for large, medium, and small cores, adopting a newly developed small core that significantly enhances performance compared to the public version. In addition to the new chip, the high-end version of the Huawei Mate 80 series will also feature the new SC595XS sensor from Sitwei, built on a domestic 22nm process, representing a top-tier domestic CMOS. This sensor boasts 50 million pixels and a large 1/1.28-inch sensor size, achieving an ultra-high dynamic range of up to 110dB.
Rumors suggest NVIDIA is considering a $73 billion acquisition of MediaTek?
According to foreign media reports, MediaTek and NVIDIA are collaborating on the development of a super chip GB10, with both parties having highly complementary technologies and a close cooperative relationship, leading to speculation that NVIDIA may consider acquiring MediaTek for $73 billion.
In response to these rumors, MediaTek stated, “This is not true.”
Industry insiders believe this is speculation, as the acquisition price is too low and would face strict scrutiny from regulatory agencies, making it unlikely for both Taiwan and mainland China to approve it. Additionally, NVIDIA has currently transformed into an AI infrastructure company, gradually moving from the chip side to the system side, making the benefits of acquiring MediaTek relatively limited for the system side.
Foreign media pointed out that GB10 combines MediaTek’s CPU and storage design capabilities with NVIDIA’s GPU design advantages. NVIDIA has demonstrated how to closely integrate its Blackwell GPU with industry-standard interfaces and agreements into efficient packaging, leveraging MediaTek’s experience in system-on-chip (SoC) design.
Industry experts noted that if NVIDIA truly intends to acquire, $73 billion may not be enough to persuade shareholders and management. In past large semiconductor industry mergers, acquisition premiums often need to reach double-digit percentages or even several times the market price. Therefore, estimating the acquisition price based on current market value would be difficult to gain internal approval from MediaTek.
Cadence acquires Hexagon’s design and engineering business for $2.25 billion.
The American EDA giant Cadence announced the acquisition of Hexagon’s design and engineering business (D&E) for €2.7 billion (approximately RMB 22.5 billion). This move is not only a significant step following its acquisition of BETA CAE in 2024 but is also seen as a key step in accelerating its expansion into the engineering simulation field beyond electronic design automation.
The transaction is expected to be completed in the first quarter of 2026, with Cadence paying 70% in cash and 30% in shares. If the deal fails to close, Cadence will incur a “reverse breakup fee” of up to €175 million.
Cadence is known for its chip design software and is a core tool supplier for major chip manufacturers such as NVIDIA and Qualcomm.
Intel: 2026 will be a crucial year for manufacturing technology, with the success or failure of the 14A process to be revealed.
According to foreign media reports, Intel believes that 2026 will be a critical year for its manufacturing technology, at which point it should be able to demonstrate whether the company is ready to advance to more advanced manufacturing processes.
Intel CFO Dave Zinsner mentioned that the company will only know by then whether it is ready to advance the so-called 14A process. Achieving this milestone has long been seen as a key part of Intel’s efforts to turn the tide. “At some point in 2026, we will have a clear understanding of the progress of this process.”
Zinsner reiterated that for financial reasons, Intel will only build relevant manufacturing capacity if it receives commitments from external customers willing to use the 14A process.
However, when Intel CEO Pat Gelsinger first articulated this position in July, analysts and investors expressed concern. Failing to advance the 14A process would mean Intel is abandoning its efforts to regain its position as a technology leader. This has also led U.S. politicians to scrutinize Intel more strictly. Former U.S. President Trump called for Gelsinger’s resignation, ultimately leading the U.S. government to exchange $8.9 billion for about 10% of Intel’s shares.
Industry Dynamics
Ant Group intensifies chip strategy
Recently, Ant Group has been continuously investing in chip companies. On August 26, it invested in the Shanghai-based ReRAM new storage startup, Xinyuan Semiconductor. Prior to that, on August 29, Ant Group’s Shanghai Yunya Enterprise Management Consulting Co., Ltd. invested in Shanghai Yezhi Chip Technology Co., Ltd., holding a 14.29% stake.
According to public information, Xinyuan Semiconductor was established in 2019 and covers AI computing and storage IP, large model acceleration solutions, high-performance/reliable system-level storage chips, and advanced process embedded storage fields, having achieved commercial shipments and deliveries in storage chips, computing and storage IP, and high-performance MCUs.
Public information shows that Yezhi Chip Technology is dedicated to the research and development of edge AI chips, focusing on a fully self-developed high-efficiency NPU core, providing AI chips and hardware-software collaborative solutions for smart glasses, mobile phones, robots, and more.
The U.S. plans to legislate to require NVIDIA to prioritize U.S. AI orders and prohibit the export of high-end GPUs.
The U.S. Senate released a preliminary version of the annual defense policy proposal on Tuesday, which includes a requirement for U.S. AI processor developers to prioritize domestic orders for high-performance AI processors before supplying overseas buyers, explicitly prohibiting the export of the highest-end AI GPUs. Legislators have dubbed this proposal the “2025 National AI Acquisition and Innovation Act” (GAIN AI Act), aimed at ensuring that U.S. “small businesses, startups, and universities” can access the latest AI GPUs from companies like AMD and NVIDIA before other countries’ customers. However, once this bill is officially enacted, it will impose a heavy burden on U.S. companies.
“Advanced AI chips are like jet engines, which will help the U.S. AI industry maintain its leading position in the next decade,” said Brad Carson, chairman of the American Responsible Innovation Association (ARI). “Currently, the supply of these chips is limited globally, meaning that for every advanced chip sold, the U.S. cannot leverage it to accelerate its own R&D and economic growth. As we compete for the leading position in this dual-use technology, incorporating the GAIN AI Act into the National Defense Authorization Act will be a significant victory for U.S. economic competitiveness and national security.”
The GAIN AI Act requires developers of AI processors like AMD or NVIDIA to prioritize U.S. buyers before selling advanced AI hardware to other countries (including allied nations such as European countries or the UK and adversaries like China). To this end, the bill proposes to implement export controls on all “advanced” GPUs exported outside the U.S. (details to follow) and to deny export licenses for the “most powerful chips.”
Rumors suggest EssilorLuxottica is considering doubling its investment in Nikon.
According to reports, the world’s largest eyewear company, EssilorLuxottica, is negotiating to increase its stake in Tokyo-based camera and optics company Nikon to “approximately 20%”.
Media reports indicate that this European eyewear giant initially invested 5.1% in Nikon in 2024, but as of today, its stake has approached 9%. Nikon has a long history in the eyewear field through its partnership established in 2000 with Essilor (before merging with Luxottica), aimed at providing optical lenses for Japanese customers. This agreement was renewed in 2024, and now both companies are considering deeper cooperation.
Representatives from both Nikon and EssilorLuxottica declined to comment. Bloomberg reported that discussions are still in the early stages, and it is uncertain whether the deal will actually occur. However, if the deal is completed, a 20% stake in Nikon would be quite substantial.
New Technology
Zhongwei Company launches six new semiconductor equipment products
On September 4, at the 13th Semiconductor Equipment and Core Components and Materials Exhibition (CSEAC 2025), Zhongwei Semiconductor Equipment (Shanghai) Co., Ltd. (hereinafter referred to as “Zhongwei Company”) announced the launch of six new semiconductor equipment products. These devices cover key processes such as plasma etching (Etch), atomic layer deposition (ALD), and epitaxy (EPI).
In terms of etching technology, Zhongwei Company released two new products that provide leading and efficient solutions for customers in the fields of ultra-high aspect ratio etching and metal etching.
Among them, the new generation ultra-high aspect ratio plasma etching “tool”—the CCP capacitive high-energy plasma etching machine Primo UD-RIE® is based on the mature Primo HD-RIE® design architecture and has been fully upgraded, equipped with six single reaction chamber reaction chambers. By utilizing a lower frequency and higher power RF bias power supply, it provides higher ion bombardment energy, meeting the stringent requirements for ultra-high aspect ratio etching while balancing etching precision and production efficiency.
The Primo Menova™ 12-inch ICP single-chamber etching equipment, which was also unveiled, focuses on the metal etching field, particularly excelling in etching metal Al lines and Al blocks, widely applicable to power semiconductors, memory devices, and advanced logic chip manufacturing, serving as a core device in wafer fabrication’s metallization process.
In this new product launch, Zhongwei Company introduced the 12-inch atomic layer deposition product Preforma Uniflash® metal gate series, which has become a highlight in the thin film deposition field. The product adopts Zhongwei’s unique dual reaction chamber design, allowing the system to flexibly configure up to five dual reaction chamber reaction chambers.
OmniVision Group releases high-voltage isolation driver chip ORD110x
OmniVision Group has developed a highly integrated high-voltage isolation gate driver chip ORD110x to meet the driving needs of IGBT/silicon carbide (SiC), supporting a maximum of 1500V for motor control systems. It employs capacitive isolation technology, supporting surge isolation voltages of up to 10,000V, with common-mode transient immunity exceeding 150V/ns, ensuring the entire control system operates safely and stably in higher voltage environments.
ORD110x features a dead time protection function, with dead time protection time settings of 0us/0.5us/1.0us/1.5us/2.0us. When the driver chip detects that the dead time between the PWM input pins INP and INN is less than the dead time set internally in the chip, the ORD110x will force the output to the chip’s internal dead time, avoiding the risk of IGBT/SiC experiencing upper and lower bridge direct conduction.
ORD110x samples the analog voltage through the AIN pin on the high-voltage side, transmits it to the low-voltage side in an encoded manner, and outputs it via the AOUT pin in a 400KHz PWM manner. Users can obtain the sampled value of the analog voltage by acquiring the duty cycle of its PWM. The input voltage range for AIN is 0.6 to 4.5V, and the output accuracy of AOUT across the full temperature range is ±4%.
Investment and Financing
Xinqing Technology completes hundreds of millions in B+ round financing
Hubei Xinqing Technology Co., Ltd. (referred to as “Xinqing Technology”) has completed a B+ round of financing, raising hundreds of millions of RMB. This round of financing was jointly participated by several well-known investment institutions, including CICC, Agricultural Bank Investment, Guozhu Capital, Wuhan Economic Development Industry Fund, TEDA Kechuang Investment, Hubei Gaotou, and Jinyu Maowu. After this round of financing, Xinqing Technology will further accelerate its R&D process in the automotive electronics chip field, enhancing its market competitiveness.
Public information shows that Xinqing Technology was established in 2018 in the Wuhan Economic and Technological Development Zone, jointly funded by Geely Holding Group’s smart technology company Zhejiang Yikatong Technology Co., Ltd. and Arm China, with R&D and sales branches in Wuhan, Beijing, Shanghai, Shenzhen, and Shenyang. It focuses on designing, developing, and selling advanced automotive electronic chips, with the mission of “allowing everyone to enjoy the joy of driving,” aiming to become a world-leading provider of automotive electronic chip solutions.
In 2021, Xinqing Technology launched the first domestically produced 7nm automotive-grade intelligent cockpit chip “Longying No. 1,” which has been applied or designated in dozens of major models both domestically and internationally, including FAW Hongqi Tian Gong series, Geely Lynk & Co series, Galaxy series, and overseas models from Volkswagen in Europe and America, with cumulative shipments exceeding one million pieces.
In 2024, Xinqing Technology will launch the all-scenario high-level assisted driving chip “Xingchen No. 1,” which adopts a 7nm automotive-grade process, directly competing with internationally advanced mainstream products, achieving comprehensive improvements in key indicators such as CPU performance, AI computing power, ISP processing capability, and NPU. Xinqing Technology has revealed that it is developing the next generation of cockpit chips “Longying No. 2 Ultra” and “Longying No. 2 Lite.”
Statement：This article is organized by Electronic Enthusiasts. Please indicate the source above when reprinting. For group discussions, please add WeChatelecfans999. For submission, exposure, and interview requests, please email [email protected].
More Hot Articles to Read
-
Goldman Sachs: China’s lithography machines lag by at least 20 years
-
China’s semiconductor technology surpasses South Korea; Dongxin Co., Ltd. plans to invest 211 million yuan in Shanghai Lishuan
-
Breaking! NVIDIA H20 no longer for sale?
-
Domestic lithography machine makes significant breakthroughs! Precision can reach 0.6nm; media “exposes” that the U.S. has implanted trackers in chip shipments
-
Shocking! BYD’s car exports surge by 130%; AI startup Perplexity plans to acquire Google’s Chrome
-
Micron Technology lays off in China; reports indicate that the prices of 8GB small-capacity storage chips for intelligent driving basic chips are starting to soar