On May 20, NVIDIA reported a robust quarterly performance, with both revenue and data center income exceeding market forecasts. The company also announced plans for a significant stock buyback and an increased dividend. During the earnings call, management conveyed an optimistic outlook. NVIDIA's CFO stated that the company is experiencing accelerating demand and has raised its forecast for total annual AI industry spending by the end of this decade (2021-2030) to $3-4 trillion. CEO Jensen Huang commented on the next-generation product line, Vera Rubin, describing its start as "good" and anticipating that the product will face supply constraints throughout its entire lifecycle. This suggests strong demand is expected to continue into the next architectural cycle. Management also indicated that the AI industry is prioritizing research, development, and investment, and the company is fully confident in achieving its $1 trillion revenue target. On specific business fronts, NVIDIA projects its CPU revenue could reach $20 billion this year and expressed confidence in securing sufficient supply to support ongoing growth.
Competitive Landscape: Inference Chip Alternatives Remain a Niche
Addressing market approaches like those from Groq, which focus on specialized inference chips, Jensen Huang was clear, stating such chips will "remain a niche for quite some time to come." This implies that NVIDIA's dominant position in the AI compute market is not expected to face substantial challenges in the near term. This statement directly addresses market concerns about whether compute demand during inference phases might shift towards lower-cost specialized chips. Regarding the capital expenditure trends of hyperscale cloud service providers, Huang noted they will continue large-scale spending, and NVIDIA's growth rate is projected to outpace this capital expenditure growth. He specifically mentioned AI startup Anthropic, stating its compute demand will be "enormous."
Comments