Does It Matter Who Leads the AI Agent Race? Wall Street Is Forming a New Consensus: Whether Meta, OpenAI, or Google Wins, Chips, Memory, and Power All Win

Stock News
Yesterday

Wall Street professionals broadly agree that in the competition for AI agents, chipmakers remain the best way for stock investors to participate in this wave 鈥?even if Meta Platforms' Muse product appears to have temporarily seized the leading position.

"Leadership is almost like a game of hot potato," said Paisley Nardini, head of investment strategy at Tema ETFs, which manages $3.8 billion in assets and holds Meta shares. "I've already heard people joking that in a few weeks we'll forget about Muse and chase the next hot thing."

So-called AI agents refer to autonomous AI systems capable of handling multi-step tasks such as bookings and financial management. The rise of such products has triggered a chain reaction in the market: investors are weighing which products may ultimately be chosen by enterprises and consumers, while also assessing the potential disruptive impact they may have on traditional business models.

Yet amid these questions, one certain trend has emerged: the new generation of AI agents requires far more computing power than previous chatbot-focused AI tools. This means chipmakers and other segments of the AI infrastructure sector will continue to benefit. Since ChatGPT's debut nearly four years ago, these areas have remained the mainstream choice for investors looking to position themselves in emerging technologies.

Computing Demand Surges as Agents Bring a "Step-Change"

The fundamental difference between AI agents and previous AI tools lies in a radical shift in their computing architecture. Advanced Micro Devices (ASX: AMD) noted in July when launching its Helios AI rack product that agentic AI is driving a "step-change" in computing demand, because agents need to reason through multiple steps, invoke tools, access data, and coordinate workloads. AMD estimates that by 2026, about 60% of global AI computing power will be used for inference tasks.

This trend is reshaping the demand structure across the entire semiconductor supply chain. According to TrendForce research data, a traditional AI data center requires about 30 million CPU cores per gigawatt of power, but in the AI agent era, that demand will surge to 120 million cores 鈥?a fourfold increase. The future CPU-to-GPU ratio is expected to shift from the current stark disparity to between 1:1 and 1:2, which will significantly boost CPU market demand.

"Rather than trying to bet on which company will ultimately build the best agent, we prefer to hold the underlying foundation on which agents run 鈥?stocks that benefit no matter which company leads," said Rick Lear, chief investment officer at Lear Investment Management. His concentrated portfolio focuses on names poised to benefit from the AI agent wave, such as NVIDIA (ASX: NVDA), Micron Technology (ASX: MU), and utility company Williams.

"Leadership in the tech industry keeps changing hands, but as long as the agent赛道 explodes overall, the market will need more chips, more memory, more power, and more infrastructure," he noted. "That underlying logic won't change, even if Meta falls from its current spotlight again."

So far this year, AI infrastructure stocks have largely dominated the market rally. Of the 10 best-performing stocks in the S&P 500, nine are related to this theme. Chipmakers have been particularly strong, with the Philadelphia Semiconductor Index surging 84%, driven by AI-related names such as Micron Technology, Marvell Technology (ASX: MRVL), Intel (ASX: INTC), and AMD.

Competition in the Agent Space Intensifies

Meta launched its personal AI assistant Muse last month, quickly garnering massive downloads and rave reviews. Muse's strong performance gave the Facebook parent a fiery September, with shares surging 27% that month 鈥?its best monthly performance in nearly four years.

OpenAI then unveiled its always-on AI agent Dots at its annual developer conference in September, running on the GPT-6 Astra platform with its own cloud computer and browser, available to ChatGPT Pro and enterprise premium users. Meanwhile, Alphabet (ASX: GOOGL)'s Gemini Spark and SpaceXAI's Grok Bot are also actively competing for AI agent market share.

Muse's rapid rise illustrates just how quickly AI market sentiment can shift, with momentum often moving swiftly to the newest or hottest product. While Meta is currently riding high, it faced heavy scrutiny in the first half of the year as the market questioned whether its tens of billions of dollars in AI spending were yielding results. At the same time, Alphabet went from being viewed as an AI leader at the start of the year to facing increasing scrutiny recently.

This sharp volatility is exactly why many investors consider AI infrastructure stocks more reliable than participants in the agent race. "If you're trying to bet on who among Meta, Alphabet, Apple (ASX: AAPL), or others will have the last laugh, right now it's about as good as a coin flip 鈥?you can't predict the outcome," said Nardini of Tema ETFs. "Positioning in infrastructure is a more direct path to profits, because the adoption of agents will drive demand for this type of hardware over the long term."

This view is rapidly becoming a market consensus, pushing analysts to continuously raise earnings expectations for AI infrastructure. Statistics show that the sector is expected to see net profit growth of 63% and revenue growth of 54% in 2027. At the end of July, market expectations were only for profit growth of less than 48% and revenue growth of 32%. Expectations for 2028 are also being continuously revised upward, indicating that Wall Street believes this trend will remain sustainable for years to come.

The continuous evolution of AI agents is precisely the core logic supporting this judgment. Citigroup analyst Atif Malik said different types of semiconductors 鈥?including processors, memory, and networking chips 鈥?could also benefit from the development of agents. In an October 6 research report, he estimated that the potential market size compound annual growth rate for CPU chips will reach 60%.

"That's why the infrastructure approach makes so much sense," Nardini said. "Meta won't carry the torch forever. Six months from now, the industry leader could be someone else. But the demand for chips won't change."

Disclaimer: Investing carries risk. This is not financial advice. The above content should not be regarded as an offer, recommendation, or solicitation on acquiring or disposing of any financial products, any associated discussions, comments, or posts by author or other users should not be considered as such either. It is solely for general information purpose only, which does not consider your own investment objectives, financial situations or needs. TTM assumes no responsibility or warranty for the accuracy and completeness of the information, investors should do their own research and may seek professional advice before investing.

Most Discussed

  1. 1
     
     
     
     
  2. 2
     
     
     
     
  3. 3
     
     
     
     
  4. 4
     
     
     
     
  5. 5
     
     
     
     
  6. 6
     
     
     
     
  7. 7
     
     
     
     
  8. 8
     
     
     
     
  9. 9
     
     
     
     
  10. 10