Nvidia (NASDAQ: NVDA) has dominated the artificial intelligence sector by supplying the most critical hardware—high-performance processors that power and train AI models. However, competition and rapid innovation cycles are reshaping the landscape.
What’s particularly interesting is Nvidia’s strategic shift toward annual hardware releases instead of its previous two-year cycle. This acceleration matters because it keeps competitors at bay and maintains consistent demand. In the latest earnings report, CEO Jensen Huang revealed that Blackwell GPU sales have exceeded expectations, with cloud GPU inventory completely sold out.
Yet Nvidia isn’t resting on its laurels. The company has already begun manufacturing the Rubin architecture—its next-generation AI supercomputer platform. This represents a major leap in capability. Rubin can reduce inference costs to one-tenth of Blackwell’s level and requires 75% fewer GPUs to train certain AI models. Such improvements directly enhance profitability for companies building AI applications, potentially triggering another wave of adoption.
China’s Market Potential Could Be a Game-Changer
One overlooked element in Nvidia’s growth story is the Chinese market. While geopolitical tensions and export restrictions have complicated matters, recent developments suggest progress. The company has ordered its manufacturing partner to mass-produce H200 chips, betting that Chinese regulators will approve large purchases. Industry reports indicate Chinese tech firms have expressed interest in acquiring hundreds of thousands of H200 units.
If this scenario plays out, the revenue injection could substantially boost Nvidia’s stock performance.
Physical AI Opens a New Frontier—And Another Stock to Watch
Beyond traditional AI training, a new trend is emerging: physical AI—systems that understand, reason, and act in the real world. Robots and autonomous vehicles are the primary beneficiaries.
This brings us to Serve Robotics (NASDAQ: SERV), a company deeply embedded in autonomous delivery and robotics. The firm operates over 2,000 autonomous delivery robots powered by Nvidia’s Jetson Orin platform, which provide Level 4 autonomy. In just one year, Serve expanded its operational fleet by 20 times through partnerships with major restaurant chains, retailers, and logistics companies.
Management projects 2025 revenue of approximately $2.5 million, yet the company commands a market valuation exceeding $1 billion. For 2026, estimates suggest revenue could reach around $25 million. If execution matches these targets, early investors could see substantial gains.
The Investment Question
Both companies represent different risk-reward profiles. Nvidia offers stability with continued growth potential through innovation cycles and emerging markets. Serve Robotics presents a higher-risk, higher-reward opportunity in the nascent physical AI sector. Success depends on whether these firms can execute their ambitious roadmaps and capture market share in their respective domains.
The AI stock narrative is far from over—the real story may just be beginning.
This page may contain third-party content, which is provided for information purposes only (not representations/warranties) and should not be considered as an endorsement of its views by Gate, nor as financial or professional advice. See Disclaimer for details.
Can AI Stocks Deliver Million-Dollar Returns? Here's Why These 2 Firms Are Positioned to Lead in 2026
The AI Chip Battle Is Heating Up
Nvidia (NASDAQ: NVDA) has dominated the artificial intelligence sector by supplying the most critical hardware—high-performance processors that power and train AI models. However, competition and rapid innovation cycles are reshaping the landscape.
What’s particularly interesting is Nvidia’s strategic shift toward annual hardware releases instead of its previous two-year cycle. This acceleration matters because it keeps competitors at bay and maintains consistent demand. In the latest earnings report, CEO Jensen Huang revealed that Blackwell GPU sales have exceeded expectations, with cloud GPU inventory completely sold out.
Yet Nvidia isn’t resting on its laurels. The company has already begun manufacturing the Rubin architecture—its next-generation AI supercomputer platform. This represents a major leap in capability. Rubin can reduce inference costs to one-tenth of Blackwell’s level and requires 75% fewer GPUs to train certain AI models. Such improvements directly enhance profitability for companies building AI applications, potentially triggering another wave of adoption.
China’s Market Potential Could Be a Game-Changer
One overlooked element in Nvidia’s growth story is the Chinese market. While geopolitical tensions and export restrictions have complicated matters, recent developments suggest progress. The company has ordered its manufacturing partner to mass-produce H200 chips, betting that Chinese regulators will approve large purchases. Industry reports indicate Chinese tech firms have expressed interest in acquiring hundreds of thousands of H200 units.
If this scenario plays out, the revenue injection could substantially boost Nvidia’s stock performance.
Physical AI Opens a New Frontier—And Another Stock to Watch
Beyond traditional AI training, a new trend is emerging: physical AI—systems that understand, reason, and act in the real world. Robots and autonomous vehicles are the primary beneficiaries.
This brings us to Serve Robotics (NASDAQ: SERV), a company deeply embedded in autonomous delivery and robotics. The firm operates over 2,000 autonomous delivery robots powered by Nvidia’s Jetson Orin platform, which provide Level 4 autonomy. In just one year, Serve expanded its operational fleet by 20 times through partnerships with major restaurant chains, retailers, and logistics companies.
Management projects 2025 revenue of approximately $2.5 million, yet the company commands a market valuation exceeding $1 billion. For 2026, estimates suggest revenue could reach around $25 million. If execution matches these targets, early investors could see substantial gains.
The Investment Question
Both companies represent different risk-reward profiles. Nvidia offers stability with continued growth potential through innovation cycles and emerging markets. Serve Robotics presents a higher-risk, higher-reward opportunity in the nascent physical AI sector. Success depends on whether these firms can execute their ambitious roadmaps and capture market share in their respective domains.
The AI stock narrative is far from over—the real story may just be beginning.