GTC 2026: Nvidia to unveil next gen AI breakthroughs to outpace rivals

CEO Jensen Huang is expected to reveal new products and partnerships, providing updates on the full-stack roadmap from the current Rubin architecture to the upcoming Feynman generation. The event will underscore advances in AI chips, data centers, AI agents and physical robotics. The primary motive is to assure investors that Nvidia's strategy of reinvesting profits back into the AI ecosystem is yielding results.

In this connection, eMarketer analyst Jacob Bourne said: “I expect Nvidia to present a full-stack roadmap update from Rubin to Feynman while emphasizing inference agentic AI, networking and AI factory infrastructure.”

Market Shift and Competition

Prominent customers like Meta and OpenAI are developing their own Application-Specific Integrated Circuits (ASICs) to reduce reliance on Nvidia's general-purpose GPUs. While analysts expect Nvidia to maintain its 90 percent market share in the near future, they predict a potential loss of shares starting in 2027 as big tech companies scale their in-house chip programs.

The return of the CPU

As agentic AI grows, the bottleneck moves to the orchestration layer, which is managed by CPUs. Nvidia plans to showcase servers that highlight its own CPU capabilities to compete with Intel and AMD.

Optical networking

Nvidia has invested $4 billion across Lumentum and Coherent to develop co-packaged optics. This uses lasers to send data between chips faster and more efficiently than traditional wiring, though scaling this technology remains a significant challenge.

The Next Era: Agentic AI

The industry is evolving and moving towards a future where fleets of AI agents carry out tasks autonomously. Because these agents will be so numerous, a new layer of AI middle managers is required to sit between the human user and their AI agents.