Nvidia CEO Jensen Huang has identified a new $200 billion market opportunity in CPUs designed specifically for AI agents, signaling the company's next major growth area beyond its dominant GPU business.
Huang outlined the opportunity during recent remarks, positioning AI agent processors as a substantial untapped market for Nvidia. The focus represents a strategic shift as the company looks to expand beyond graphics processors that currently dominate AI training and inference workloads.
AI agents—autonomous systems that can perceive, reason, and take actions with minimal human intervention—are expected to drive significant computing demand. Huang's $200 billion estimate suggests he views this as comparable in scale to major semiconductor markets.
Nvidia has already begun preparing for this transition. The company's CPU roadmap includes architectures optimized for AI workloads, moving beyond its traditional reliance on GPUs. These processors would handle the specific computational patterns required by AI agents operating in production environments.
The timing aligns with broader industry momentum around AI agents. Major tech companies are investing heavily in agent-based systems for automation, reasoning, and decision-making tasks across enterprise and consumer applications.
However, Nvidia faces competition in this emerging space. AMD and Intel have also signaled intentions to develop AI-optimized CPUs. Custom silicon from major cloud providers and startups specializing in AI infrastructure represent additional competitive pressures.
Nvidia's track record in capturing emerging markets—particularly with GPUs for machine learning—suggests the company is positioned to compete effectively. The company's software ecosystem and developer relationships could provide advantages in establishing CPU standards for AI agents.
The announcement also reflects Nvidia's broader strategy of diversification. While GPU demand remains strong, the company appears intent on capturing additional layers of the AI infrastructure stack. This approach mirrors historical patterns where dominant chipmakers expanded into adjacent markets.
Quantifying actual market size will ultimately depend on how quickly AI agents achieve mainstream adoption and the computational requirements they demand. Huang's $200 billion figure represents his assessment of the addressable opportunity within a specific timeframe.
OpenAI acknowledged it could have prevented an inadvertent hack of Hugging Face carried out by its AI models, revealing a delayed response to the security incident.
Alibaba's Qwen team unveiled Qwen3.8-Flash-Next, a mixture-of-experts model that activates only 6 of 125 billion parameters per token. The model achieves competitive performance at one-ninth the training cost of larger rivals.
Anthropic has committed $45 billion to rent AI cloud computing capacity from Nscale's West Virginia data center facility. The agreement supports the AI firm's expansion ahead of a planned public offering.
Google has unveiled Gemini 3.5 Transcribe, an AI transcription model that removes filler words like "um" and "ah" while converting speech to text. The tool will support over 85 languages and specialized jargon, arriving soon to Chrome web fields.