Muse is a hit, but how will the AI industry chain change? A visit to 5 hard-tech companies.

With Muse gaining popularity, what changes will occur in the AI industry chain?

On September 8, Meta launched Muse, a personal AI assistant that can help users send emails, book trips, complete shopping, and perform other tasks.

In less than two weeks, Muse topped the US App Store's free app chart. On September 21, Meta's stock price surged 11.4%.

The Muse phenomenon reignited investor optimism about chip demand. On the same day Meta surged, AMD's stock rose 10%, pushing its market capitalization past $1 trillion for the first time. CPU giant Intel soared 12%, becoming the best-performing stock in the S&P 500.

This market rally raises a question more worthy of consideration than stock prices: When AI begins to perform tasks for humans, what changes will occur in the chips, data transmission, storage, and terminal devices that support its continuous operation?

The answer won't just lie in the next wildly popular app. Models need to run, computing power needs to be supplied; cluster sizes need to expand, data transmission between chips needs to be faster; tasks are moving to phones, cars, and robots, computing needs to reach more terminals.

The changing demands for AI are spreading along the industry chain to more links.

From October 15th to 16th, Wall Street Insights will visit five AI hard technology companies to observe how these changes are being implemented in specific businesses from five perspectives: GPU and software ecosystem, storage, optical interconnect, computing power services, and edge AI.

This research trip includes visits to Muxi Technology, Xizhi Technology, Dianguang Technology, Dongxin Semiconductor, and Black Sesame Technologies. We aim to address the same question on-site: How are AI demands changing, and how will this impact companies across the industry chain?

Muxi Technology – one of the “Four Little Dragons” of domestic GPUs, continuously improves its domestic computing power layout from chips to software ecosystem around general-purpose GPUs.

Xizhi Technology – Focusing on optical interconnects and optical computing. As AI clusters continue to expand, how to transmit data more efficiently between chips and reduce energy consumption is becoming an important issue for computing infrastructure.

Electro-Optics Technology – Extending from mining equipment to AI computing services, attempting to open up a second growth curve outside of traditional manufacturing.

Dongxin Semiconductor – a domestic memory chip design company, is advancing an integrated layout of "storage, computing, and connectivity" based on its memory business.

Black Sesame Technologies – Starting with intelligent driving chips and computing platforms, we are continuously expanding into edge AI scenarios such as robots and smart terminals.

The five companies are located in different positions, which provides five perspectives for observing the AI industry chain.

Alongside businesses, there are two other investment perspectives.

Another part of this research trip involved two professional investors with extensive experience in technology investment, who each shared their insights on specific topics.

In the early hours of the morning , the head of the TMT group in the research department of Harmony Capital, a private equity firm with assets exceeding 10 billion yuan , and the investment manager of the TMT Innovation Selection Fund, has been tracking AI industry and technology investment for a long time.

In December 2024, he systematically discussed the evolution of the AI industry from computing infrastructure to application commercialization in the "Finding the Next Nvidia" masterclass on Wall Street Insights.

At the time, he focused on a key shift: as AI moves from assisting humans to performing tasks independently, the reasons why companies pay for AI will also change. Understanding what tasks an agent can accomplish and how much value it creates is a crucial clue to understanding the commercialization of AI.

In the early morning, he summarized his research framework into three scales: in the long term, technological progress and industrial space; in the medium term, commercialization progress; and in the short term, catalysts and market expectations.

During this research trip, he will also share his insights on the latest industry changes, and reintegrate his on-site observations of companies into the investment research framework.

Wei Haitao is the founder of Shanghai Shengyu Qiansheng Investment Management Co., Ltd., and has been engaged in financial investment and fund management for over 20 years. He previously served as a partner at Shengyu Investment, focusing on venture capital, equity investment, and private placement investment in listed companies. He has deep experience in the semiconductor and intelligent equipment industries and has led the investment and industrial empowerment of dozens of science and technology innovation projects.

With extensive experience in primary market investment, he served as a director of Tunan Technology from 2015 to 2021, and witnessed the company's listing on the ChiNext board in 2020. His former employer, Shengyu Investment, also had a long history of investing in hard technology companies such as Huatian Technology, Tunan Technology, and Huahai Chengke.

His long-term investment in the primary market has led him to focus on another set of questions when observing technology companies: how a technology can truly be industrialized, how a company can move from product to market, and how to build long-term competitiveness amidst industry changes.

In this research, Wei Haitao will also share his methods for observing and researching hard technology companies, drawing on his more than 20 years of experience in technology industry investment.

One has been tracking the changes in the AI industry from the secondary market for a long time, while the other has been seeking out and supporting the growth of hard technology companies in the primary market.

Five companies, five industry segments. Two different investment perspectives.

Over the course of two days, we hope to put the AI information that is usually scattered in news, announcements, industry conferences, and market discussions back into a more complete industry map: to see where the technology is heading, what companies are doing, and how professional investors understand these changes.

If you've already read a lot of AI news and want to get closer to the industry scene—

From October 15th to 16th, we visited enterprises together.

Risk warning and disclaimerInvesting involves risk; please exercise caution. This article does not constitute personal investment advice and does not take into account the specific investment objectives, financial situation, or needs of individual users. Users should consider whether any opinions, views, or conclusions in this article are suitable for their specific circumstances. Any investment decisions made based on this information are at your own risk.