NVIDIA enters the AI agent era, accelerating the arrival of the high-growth concept stocks

Model iterations are accelerating the arrival of the intelligent agent era, which will profoundly impact AI application ecosystems across all scenarios, including office work, development, and content creation.

HuiLv Ecology’s Largest Two-Year Increase Exceeds 13 Times

Today (March 10), the early trading session saw a significant rebound in A-share technology sectors, with leading gains in CPO, PCB, optical communication, advanced packaging, and high-speed copper cable connections.

The CPO concept performed the best, with stocks like Jinxuexing hitting a 20-cent limit-up; HuiLv Ecology (001267) also reached a 10-cent limit-up, with its latest share price at 45.68 yuan, setting a new record high. According to Securities Times Data Treasure, compared to the recent two-year low, HuiLv Ecology’s maximum increase has exceeded 13 times.

The company’s previous earnings forecast indicated that net profit in 2025 is expected to be between 86 million and 110 million yuan, representing a growth of 0.27% to 28.26% over the same period after restructuring. During the reporting period, Wuhan Junheng was officially included in the company’s consolidated financial statements from February 2025. The company’s stake in Wuhan Junheng increased from 35% to 51%, which contributed to a corresponding rise in net profit attributable to shareholders.

According to research information disclosed by HuiLv Ecology in November 2025, Wuhan Junheng is currently focused on expanding capacity, developing customers, and ensuring product quality. Regarding capacity expansion, Wuhan Junheng’s Phase 1 new production line in Ezhou has recently gone into operation, with an additional annual capacity of 1.5 million units, mainly producing 400G and 800G high-speed optical modules. The Phase 2 project will add 3 million units per year and is expected to be officially operational by 2027.

NVIDIA Plans to Launch an Open-Source AI Intelligent Agent Platform

On March 9 (local time), media reports cited insiders revealing that NVIDIA plans to launch an open-source AI intelligent agent platform called NemoClaw. The platform aims to enable enterprises to deploy AI agents within their workflows to replace employees in executing specific tasks.

Sources said that regardless of whether these companies’ products run on NVIDIA chips, they will be able to access the platform.

NVIDIA has begun seeking cooperation with several major software companies, including Salesforce, Cisco, Google, Adobe, and CrowdStrike, with plans to officially unveil the product around next week’s annual developer conference in San Jose.

Recently, an open-source AI agent framework called OpenClaw has become popular worldwide. Due to its logo featuring a red lobster, deploying and using OpenClaw is also called “Lobster Farming.”

At the Morgan Stanley Technology Conference on March 6, Jensen Huang highly praised OpenClaw, calling it “the most important software release of this era.” He emphasized that the Linux operating system, which took thirty years to permeate the infrastructure of global servers, mobile phones, and supercomputers, has achieved widespread adoption, and OpenClaw has surpassed this scale in just three weeks.

The Era of AI Intelligent Agents Begins

Industry insiders explain that AI intelligent agents are AI assistants capable of a closed-loop “thinking—acting—reflecting” process. Unlike ChatGPT’s one-question-one-answer mode, AI agents can understand complex goals, autonomously decompose tasks, invoke tools to execute, and continuously optimize strategies during actions.

Since 2026, China’s AI field has seen a “new wave” of large models—companies like Alibaba, ByteDance, Kimi, DeepSeek, and Zhipu have launched new-generation large models. These model providers are no longer solely focused on parameter scale but are also developing native agent capabilities, with AI technology evolving from “chatting” to “getting things done.”

From the emergence of the “system-level AI assistant” concept to widespread discussions about “AI taking over phones,” a series of landmark events signal that China is making significant strides into the “AI intelligent agent era.”

On a broader scale, this year, intelligent agents have penetrated various industries, with over 30 million actively functioning in different sectors. IDC reports that the number of active AI agents in Chinese enterprises is expected to surpass 350 million by 2031, with a compound annual growth rate of over 135%. In this rapid development, issues such as digital property rights and responsibility delineation are emerging in the regulatory vacuum.

湘财证券 believes that model iterations are accelerating the arrival of the intelligent agent era, which will deeply influence AI application ecosystems across all scenarios, including office work, development, and content.

According to a report by China Business Industry Research Institute, the global AI intelligent agent market size is estimated at about $5.1 billion in 2024. Analysts forecast that by 2030, the global market will approach $50 billion. Behind this rapid market expansion is the industry’s push to turn “AI from talking to doing” into practical applications.

28 Concept Stocks Show Significant Performance Growth

Data Treasure statistics show that, based on disclosed annual reports, performance briefs, and profit forecasts (using the lower limit if available), 28 AI intelligent agent concept stocks are expected to see net profit growth of over 20% in 2025 (including turning losses into profits). Stocks expected to turn profitable include Jincai Hulian, Xin’an Century, Yinli Media, 360, Winsight, DeAn Diagnostics, BlueFocus, and Lio股份.

Taking 360 as an example, the company expects net profit between 213 million and 318 million yuan in 2025, turning from loss to profit. The improved performance is mainly due to increased investment income from associated companies. In the first half of 2025, the company launched China’s first “Super Search Intelligent Agent”—NanoAI Super Search Intelligent Agent, marking the entry of AI search into the 3.0 era.

Besides stocks expected to turn profitable, others with significant profit growth include DianSheng股份, GuangTing Information, Advanced Digital, East Tuo Technology, and QingMu Technology.

According to Data Treasure, as of March 9, among these 28 high-performing AI concept stocks, 13 have retreated more than 20% from their 2026 high points. QingMu Technology, Yinli Media, DeAn Diagnostics, BlueFocus, and Tonghuashun are among the most retraced.

QingMu Technology leads with a 40.53% decline. The company has already invested in data analysis and intelligent customer service systems using AI intelligent agents; it expects net profit between 118 million and 136 million yuan in 2025, a year-on-year increase of 30% to 50%.

(Article source: Data Treasure)