On Monday, Micron Technology (MU) was the most heavily traded US stock, closing down 3.88% with $23.47 billion in turnover.

On Monday, South Korean memory giants SK Hynix and Samsung Electronics were under pressure throughout the Korean trading session, and the negative market sentiment spread further to US stocks. Market analysis suggests that heavy foreign selling may be a key reason for the decline. Data from the Korea Exchange shows that as of October 7, foreign investors' stake in Samsung Electronics fell to 46.37%, the lowest since July 16, 2009, a roughly 17-year low. The market has developed new concerns about the global competitive landscape in the memory chip sector, and Micron Technology faces competitive pressure from rivals such as Samsung and SK Hynix. ChangXin Memory Technologies President Cao Kanyu announced at an industry event over the weekend that the company has made key progress on a next-generation 4F² memory cell architecture and plans to launch server memory products based on this architecture before the end of 2026. This design uses proprietary vertical surrounding gate transistors to suppress leakage, allowing more memory cells to be packed onto a single chip, directly targeting the high-margin DRAM used in AI servers.

No. 3 was Meta Platforms, Inc. (META), which closed up 3.38% with $18.479 billion in turnover. Meta's personal AI assistant Muse retains users at a level close to ChatGPT. According to Silicon Valley venture capitalist Deedy Das, citing Sensor Tower data on X, Muse's 7-day return visit rate exceeds 30%, essentially on par with ChatGPT; its 14-day retention rate is 22%, slightly below ChatGPT. In addition, Meta released Muse Wonder and the more powerful Wonder Full AI models. According to test data published by Meta, the two models have reached levels close to Fable 5.1, Opus 5.5 and GPT-6 Astra in multiple capability tests, and even slightly outperformed them in some agent and knowledge work benchmarks. At Monday's promotional event, Zuckerberg hinted that Meta will further increase investment, saying the company "is already working toward the next level." He also said the pace of model development "may feel noticeably faster" going forward.

No. 4 was NVIDIA (NVDA), which closed down 1.29% with $15.866 billion in turnover. NVIDIA and Cisco jointly launched a new pluggable optical module form factor, USFP (Ultra-Dense Small Form Factor Pluggable), aimed at expanding the network scale of AI factories. USFP is the first module to support 400Gb/s electrical signaling per channel, with a single-module capacity of up to 12.8Tb/s, enabling AI builders to connect more computing systems without adding network layers while reducing transmission power consumption per bit of data. Compared with existing 200Gb/s-per-lane optical modules, USFP improves bandwidth density by 4x and energy efficiency by 1.3x. NVIDIA and Cisco are working with TE Connectivity and Amphenol to build a multi-source hardware ecosystem around USFP.

No. 6 was SanDisk, which closed down 5.97% with $12.933 billion in turnover.

No. 7 was SpaceX, which closed up 0.52% with $11.007 billion in turnover. Morgan Stanley analyst Adam Jonas said SpaceX's upcoming 15th Starship test flight could become an important step toward full rocket reusability and provide support for the company's satellite communications and artificial intelligence business expansion. The mission may attempt to use the launch tower's mechanical arm to catch the returning Starship vehicle; if progress is made, it could help lower launch costs and accelerate Starlink expansion. Morgan Stanley maintained its "Overweight" rating and $300 price target on SpaceX, while flagging risks such as technical failures during the test flight, development delays and rising infrastructure spending.

No. 8 was Microsoft, which closed up 1.14% with $9.884 billion in turnover. Microsoft announced on Monday that its next-generation Maia 300 chip will begin deployment in data centers next year, designed specifically for AI inference tasks. Microsoft expects Maia 300 to process 40% more tokens per megawatt than mainstream GPU systems, meaning it can handle more AI workloads at the same energy consumption.

No. 9 was AMD, which closed down 2.28% with $9.083 billion in turnover.

No. 10 was Apple, which closed up 1.29% with $8.802 billion in turnover. Apple announced that it will hold a special "Welcome Home" event on Tuesday (October 13), and is expected to launch new smart home products — this is also Apple's latest update in the smart home field after John Ternus became CEO last month. It is reported that the devices released this time will include a new generation Apple TV 4K, HomePod mini, and a new home hub device.

No. 11 was Tesla, which closed down 1.22% with $7.648 billion in turnover.

No. 12 was Intel, which closed down 3.20% with $7.481 billion in turnover. Intel CEO Chen Liwu said that Chinese frontier AI models have lower token costs and have been spreading very quickly. Chen Liwu said that the performance of Chinese AI models such as DeepSeek, Kimi and GLM is already good enough, and he firmly supports open source, which is the optimal path to improving efficiency. He also believes that the market will determine the application scenarios for different models.

No. 14 was Amazon, which closed down 1.11% with $6.906 billion in turnover.

No. 15 was Google A, which closed up 0.30% with $6.721 billion in turnover. (Screenshot from Sina Finance APP — Quotes — US Stocks — Stock Market Section; swipe left for more data) Download the Sina Finance APP