NVIDIA (NVDA) GTC Taipei Financial Analyst Q&A summary

Key announcements and financial strategy

• Announced plans to return 50% or more of free cash flow to shareholders this year, next year, and beyond, with increased stock repurchases and dividends over time.

• Emphasized a recently announced $80 billion share repurchase program and a 25x increase in dividend.

• Company resegmented financial reporting to provide greater transparency into business operations, highlighting three main segments: hyperscalers, AI clouds/enterprise, and robotics edge.

• Noted robust supply chain support and ability to sustain high growth, despite global supply constraints.

• Highlighted the growing software license business, projecting billions in annual revenue from AI enterprise software.

Technology and product innovation

• Introduced Vera Rubin, a new CPU architecture designed for agentic computing, with industry-leading bandwidth and custom ARM cores.

• Vera Rubin and Grace Blackwell platforms enable distributed, disaggregated AI workloads across cloud, PC, workstations, and edge devices.

• NVLink 72 and NVLink Scale-Up Switch enable lowest-cost token generation and position the company as a leading AI networking provider.

• Announced partnerships with major software vendors to accelerate applications for agentic AI, including Adobe, Cadence, Synopsys, Ansys, Dassault, and Siemens.

• Spectrum-6, the first 800 Gigabit CPO, is designed to scale AI factories to hundreds of thousands or millions of GPUs.

Market outlook and business model

• Projected significant growth in CPU market, with Vera CPUs expected to be used in all NVIDIA GPU data centers and beyond.

• Stressed that the agentic computing market is a new, rapidly expanding segment, with demand for CPUs and GPUs driven by AI agents.

• Emphasized a balanced approach to open models on devices and large models in the cloud, supporting both for future AI workloads.

• Highlighted the importance of performance per watt and energy efficiency as key differentiators for AI data centers.

• Asserted that AI infrastructure is highly resilient to economic downturns due to its role in driving productivity and innovation.