chips

According to The Information, OpenAI plans to purchase Google's artificial intelligence chips.

The Shanghai Municipal Economic and Information Technology Commission issued the "Action Plan for Accelerating the Development of a New Generation of Intelligent Manufacturing in Shanghai." The plan proposes to strengthen the industrial network and computing power infrastructure, support manufacturing enterprises in carrying out new industrial network transformations, promote the integrated evolution of control systems, industrial networks, and computing power facilities, and advance the pilot projects for industrial 5G/EUHT independent private networks.The plan also proposes to support the deployment of high-quality networks such as 5G-A and 10 Gigabit optical networks, accelerate breakthroughs in AI high-performance chips, and expedite the construction of a "hub-region-edge" computing power facility system. At the same time, it aims to strengthen the overall scheduling of intelligent computing resources, accelerate the construction of national computing power interconnection nodes, promote standardized access and on-demand invocation of computing power resources, and advance the inclusive supply actions of computing power services, enhancing the efficiency of the industrial intelligent computing cloud platform.

Recently, there has been discussion in the market about Cambrian's chip shipments exceeding 1 million units next year. At the Cambrian third quarter earnings briefing on the afternoon of October 26, Cambrian's Chairman and General Manager Chen Tian Shi responded that the company has never engaged in any related communication.Chen Tian Shi stated that market information is complex, and he urged investors not to trust information from unknown or unverified sources. The above response was reported by Yicai.

Most Hong Kong chip stocks fell, with Hua Hong Semiconductor (01347.HK) down over 7%, ASMPT (00522.HK) down over 6%, Wallen Technology (06082.HK), Tianshu Zhixin (09903.HK), and Zhaoyi Innovation (03986.HK) down over 5%.Some artificial intelligence stocks strengthened, with Qunhe Technology (00068.HK) up over 15%, Chipone Technology (02166.HK) up over 4%, NetEase-S (09999.HK) up over 3%, and Meitu (01357.HK) up over 2%.

According to North Korean Commerce reported on October 25, after a sharp decline in the South Korean stock market, individual investors have re-entered to buy leveraged products betting on the rebound of Samsung Electronics and SK Hynix. Data from the Korea Exchange and Koscom ETF Check shows that from October 1 to 23, individual net purchases of 14 related products amounted to 315.4 billion Korean won, of which 7 related to SK Hynix accounted for 226.3 billion won, and 7 related to Samsung Electronics accounted for 89.1 billion won.The aforementioned 14 products had a net purchase of 27.728 trillion won in July, but turned to net selling after regulatory tightening such as the increase in depositary receipts on July 31. In August and September, there were net sales of 648.4 billion won and 497.9 billion won, respectively, and this month individuals have first turned back to net buying. During the same period, KODEX SK Hynix leveraged product saw a net purchase of 15.294 billion won, ranking 11th among ETFs listed in South Korea; individuals also net bought KODEX Leverage 10.448 trillion won and KODEX Semiconductor Leverage 256.4 billion won. The South Korean Composite Index fell by 18.83% in October.Data from the Korea Securities Depository shows that from October 1 to 22, South Korean investors net bought U.S. stocks worth 4.8009 billion dollars, approximately 63.621 trillion won. Among them, the net purchase of SOXL, which tracks the daily returns of the U.S. semiconductor index at 3 times, was 3.54764 billion dollars, about 47.041 trillion won, accounting for 73.9% of the net purchases of U.S. stocks during the same period. The article also noted that leveraged products track daily returns by multiples, and when the market reverses, losses can also be magnified by 2 or 3 times, and during repeated sharp rises and falls, negative compounding effects may also occur.

Bitwise released a report indicating that Bitcoin is currently holding resiliently around $82,000, a price level close to the spot ETF holding cost. Key support levels below include the short-term holder cost price of approximately $74,800 and the real market average of about $77,600. The former can serve as a reference for judging the short-term market direction, while the latter corresponds to the price level at which half of the invested capital in the market is in profit or loss. The price range formed by these two points may become the dividing line between bulls and bears.The trading range for Bitcoin over the past 30 days and 60 days has narrowed to 1.8% and 3.0%, respectively, with the implied volatility at-the-money for each term also approaching the lower end of the distribution range over the past three years. Among them, the 3-month and 6-month implied volatility are at the 0.8 and 0.2 percentiles of historical distribution, indicating that the market's pricing for future volatility has dropped to extremely low levels.So far this week, net inflows into U.S. Bitcoin spot ETFs have reached $1.074 billion, rising in sync with the cumulative trading volume of the spot market, providing support for Bitcoin prices and offsetting some macro pressures. However, realized profits in the market have also increased, particularly among long-term holders. Meanwhile, the net position change indicator for long-term holders is negative, suggesting they may be in a distribution phase, which could be one reason for the ongoing consolidation of Bitcoin prices.

Cryptocurrency market analyst Murphy stated that BTC has accumulated approximately 921,000 chips around the $83,000 mark, approaching the milestone of 1 million. The last time a similar scale was observed was on August 3 of this year, when the chips at that price level reached about 1.05 million.Regarding the historical Coinbase wallet chips that may be included, Murphy indicated that about 550,000 BTC are related to the Coinbase wallet consolidation from a year ago, but it is currently unclear how many positions remain. According to his estimates, there may only be 100,000 to 200,000 left, so this factor can be temporarily ignored. BTC has entered the first chip gap area below $83,000. Since the price did not fully exchange hands when passing through this area previously, the price returning to this range creates an attraction effect, prompting chips to be re-exchanged and fill the gap.BTC faced resistance when attempting to move downward and may look for a direction with less resistance again. The $83,000 mark is also an important focus in the current long-short battle; any party gaining an advantage could drive the market to form a new trend.

According to Sifted, the Dutch startup Photon Bridge announced the completion of a €10 million seed round financing. The company is developing laser technology that can be used for internal connections of AI chips in data centers, aiming to enhance inter-chip connectivity performance.The company claims that its laser technology can be manufactured at scale to meet the rapidly growing demand for AI infrastructure.

According to a report from ZDNet Korea on October 20, the AI chip company Groq, which has been effectively acquired by NVIDIA, is advancing the introduction of DRAM 3D stacked packaging in its next-generation chips to replace the fast but lower-capacity SRAM solution. Groq was founded in 2016 by former Google TPU designer Jonathan Ross, and it was reported that the company was acquired by NVIDIA in a roundabout manner last year.The report states that Groq is evaluating a solution called 3DS, which vertically connects DRAM chips through silicon vias for high-capacity packaging. 3D packaging allows multiple bare chips to be stacked vertically and operate as a single chip, reducing area and improving power efficiency compared to planar solutions. Both 3DS and HBM use silicon vias, but HBM focuses on transmission speed, while 3DS maintains existing DRAM interfaces and focuses on capacity expansion. Anonymous industry insiders indicated that Groq plans to adopt 3D packaging in its next-generation chips, stacking DRAM instead of SRAM; another individual noted that market demand is shifting towards larger capacity chips, prompting Groq to choose DRAM, which has a capacity advantage.

According to a report by Reuters cited by Phoenix Technology, Qualcomm CEO Cristiano Amon informed employees in an internal memo that the company will initiate performance-focused global layoffs following the annual employee evaluations, with notifications set to begin in November. The memo reviewed by Reuters was sent out on Monday, and Amon stated that this move aims to establish a stronger performance culture and address the company's ongoing failure to meet expectations.The memo did not disclose the number of layoffs, and Reuters is currently unable to determine the scale. Amon mentioned that the layoffs will take into account business priorities, organizational needs, skills, and expertise; as the business expands, the required capabilities and skills are also changing, and it is essential to have the right people in the right positions. Qualcomm has not yet responded to Reuters' request for comment. As of September 2025, Qualcomm has approximately 52,000 employees worldwide, most of whom are full-time, with the fiscal year typically ending in late September.As one of the world's largest smartphone chip manufacturers, Qualcomm has been impacted this year by sluggish sales of consumer electronics, and soaring memory chip prices have suppressed demand in key markets. Since the beginning of the year, Qualcomm's stock price has increased by about 1.5%, while the S&P 500 index has risen by over 12% during the same period, and the Philadelphia Semiconductor Index has seen an increase of nearly 60%. On Tuesday, Qualcomm's stock price fell by about 1% in pre-market trading. In the most recent earnings report released in late July, Qualcomm projected that quarterly profits would fall below market expectations and indicated that revenue from Apple products would decline faster than previously anticipated, while stating that by fiscal year 2027, growth in data centers and other non-mobile businesses would be sufficient to offset the aforementioned losses and provide additional growth.

Elon Musk stated that the goal of the Terafab project at Gibbons Creek is to achieve mass production of chips around the third quarter of 2028, with the first test chip expected to operate normally in May or June 2027.

According to a report by Reuters, on October 20, ByteDance, Alibaba, and Tencent have ordered over 3 million RTX PRO 5500 graphics cards from NVIDIA, with a unit price of approximately $11,400, totaling over $35 billion. Due to U.S. export controls, this chip does not include the advanced Blackwell architecture and is currently the most powerful compliant NVIDIA chip available for purchase in China.Insiders have indicated that due to demand far exceeding production capacity, NVIDIA has informed partners that it may only be able to fulfill 5% to 10% of orders. Chinese tech companies primarily use this chip for AI inference and post-training model generation. Although its performance is far inferior to the banned B300 and H200 chips, in the context of ongoing computing power shortages and restrictions, the RTX PRO 5500 has become an attractive alternative. Additionally, NVIDIA plans to shift most of TSMC's capacity to the new generation Rubin products starting in 2027. Currently, ByteDance, Alibaba, Tencent, and NVIDIA have not commented on this.

According to The Information, DeepSeek is collaborating closely with Huawei to jointly design the next generation of AI chips, aiming to build models that do not rely on NVIDIA's computing power. Insiders revealed that Huawei engineers have entered DeepSeek's office to understand its model architecture and algorithm requirements, and to test how the chips handle data transmission and compression, with the first compatible chip targeting Ascend 950DT.The report pointed out that this move comes as China pushes to reduce its dependence on American chips. Limited by TSMC's foundry services and U.S. export controls, Huawei's Ascend 950DT is planned for release in 2027, but mass production may be delayed until 2028. To address the chip shortage, DeepSeek is using the TileLang programming language to bypass CUDA restrictions, while still attempting to acquire the banned NVIDIA Blackwell chips through third-party data centers. In addition, DeepSeek is forming a chip team and has begun testing clusters with over 1,000 processors.

Financial Associated Press, October 20 - Amazon launched the open-source tool TorchNeuron on the 19th local time, aiming to lower the barrier for developers to transition from NVIDIA chips to its self-developed chips, challenging NVIDIA's software barrier built on CUDA. According to reports, this tool allows AI code written in PyTorch to run on Amazon's self-developed AI chips without the need for rewriting. PyTorch is maintained primarily by Meta and is described as an open-source framework used by most AI developers.Amazon's chip design department, Annapurna Labs, introduced that with TorchNeuron, developers only need to modify one configuration to run the same code on the self-developed Trainium chip. Amazon announced the launch of this tool on the 19th local time and plans to publicly release it in a free open-source format at the PyTorch conference in San Jose, California on the 20th.Colin Brace, Vice President of Software at Annapurna Labs, stated in an interview that the launch of TorchNeuron is intended to reduce the steps required for developers to try Amazon's self-developed chips, starting from the customers' existing workflows. He revealed that over 30 early customers are already using it, covering fields such as robotics, world models, and voice AI, and mentioned that switching to Amazon chips can save costs.

At the 2026 RISC-V China Summit, Chief Scientist Meng Jianyi of Alibaba Damo Academy stated: So far, the Xuantie chips have steadily shipped over 6 billion units. Xuantie is a series of RISC-V processors launched by Alibaba Damo Academy.At the conference, Alibaba Damo Academy also released the first automotive-grade enhanced E908A chip. Meng Jianyi cited a prediction from Omida, stating that by 2030, the annual shipment volume of global RISC-V processors will grow by nearly 30% each year, and by 2030, the global market share of processor shipments will reach approximately 25%.

According to the Industrial and Commercial Times, Japan's Ministry of Economy, Trade and Industry has approved TSMC's first phase plan for the production of advanced semiconductors using 3-nanometer process at its second factory in Kumamoto Prefecture. This will be the first time a manufacturer in Japan produces 3-nanometer chips. TSMC can further expand its production capacity based on additional support from the Japanese government. Kumamoto Prefectural Governor Seiji Kimura stated that this plan will drive a new wave of creative revitalization.The Nikkei reported that the Ministry of Economy has listed the Kumamoto second factory as a recipient of subsidies, with plans to produce 3-nanometer chips, expected to officially start production in December 2028, with final capacity depending on the amount of subsidies. The Ministry will include the budget needed for subsidies in the fiscal year 2027 budget request, with the Japanese government originally planning to provide 732 billion yen in subsidies. The Asahi Shimbun pointed out that the Kumamoto second factory is still under construction and is adjacent to the already operational Kumamoto first factory.Previously, TSMC planned for the second factory to manufacture 6-nanometer and 12-nanometer processes, with a monthly capacity of 48,000 wafers, and planned to produce 40-nanometer chips; in February this year, the plan was changed to 3-nanometer, reducing the monthly capacity to 6,000 wafers. The adjustment plan for the Kumamoto first factory was also approved by the Ministry, with the original plan to produce 55,000 wafers per month covering 22 to 28 nanometers and 12 to 16 nanometers, while the new plan stops production of 12 to 16 nanometers, changing to 45,000 wafers per month for 22 to 28 nanometers.

Samsung Electronics will adopt TSMC's 3-nanometer process to produce the next generation of custom HBM chip bare dies for NVIDIA. This is the first time Samsung has collaborated with TSMC in the HBM manufacturing stage.NVIDIA has requested Samsung, SK Hynix, and Micron Technology to use TSMC's process to manufacture NVHBM bare dies and to utilize sixth-generation DRAM technology. Samsung plans to continue independently producing standard HBM and custom products for other customers.

The white-haired stock god Serenity stated that the storage chip industry is experiencing unprecedented long-term demand visibility, with large cloud service providers like Google even seeking to sign long-term supply agreements extending to 2035. However, the market is still affected by concerns over the sustainability of AI infrastructure investments and financing risks.Serenity also cited the views of Dell CEO Michael Dell, who indicated that the shortage of storage chip supply may last for about 4 years, as the construction cycle for new fabs is relatively long, and the related capacity is expected to be released by 2029 to 2030. Dell anticipates that the imbalance between supply and demand for storage chips may further worsen by 2027.