en.Wedoany.com Reported - Nvidia's single-quarter revenue is approaching the $100 billion mark, and company management stated in a recent closed-door investor meeting that revenue growth momentum is accelerating and spreading to a broader range of industries. CEO Jensen Huang and CFO Colette Kress have divided the future growth landscape into three pillars, each with its own unique drivers.
The first pillar is the AI lab sector, which currently accounts for about 20% of Nvidia's total demand. A landmark shift is that a well-known institution, which previously relied heavily on the ASIC roadmap to develop cutting-edge models, has now migrated nearly half of its computing load to Nvidia's architecture, whereas Nvidia's involvement with this client was previously almost zero. Meanwhile, most other top-tier models remain based on the Nvidia ecosystem, solidifying its position in scientific research and frontier exploration.
The second pillar comes from traditional hyperscale cloud service providers, contributing approximately half of Nvidia's revenue. Despite constraints from data center space and power supply, this sector maintains an exceptionally strong growth curve. Through the synergistic expansion of networking equipment and CPU products, Nvidia is broadening its addressable market and reaffirming its goal of reaching $20 billion in CPU business revenue by 2026. In response to competition from cloud giants developing custom ASICs, Nvidia's defensive moat lies in its advantage of the lowest token generation cost, allowing it to retain an overwhelming market share even as competition intensifies.
The third engine comes from AI cloud, industrial applications, and enterprise markets, which together account for the other half of data center revenue. Constrained by space and power supply, coupled with complex factors such as geopolitics and supply chain reshoring, sovereign AI and enterprise applications are experiencing explosive growth, and Nvidia faces almost no competition in this field. Additionally, the company has introduced a business model for new cloud clients that includes co-investment and profit sharing, which could potentially transform Nvidia into a co-owner of large-scale GPU cloud networks, deeply participating in the operational revenue of computing infrastructure.
On the product roadmap, Nvidia clarified a market rumor that the Rubin Ultra chip would be delayed until 2028, confirming that the chip will ship on schedule in 2027. The company stated that the product's physical form will be adjusted to support larger single-rack expansion, and key technology nodes such as the 800-volt power architecture and optical interconnects between racks are progressing as planned.
On the supply chain front, Nvidia warned that storage shortages could persist for several years but emphasized a dynamic balance among computing, networking, and memory—when memory supply is under pressure, it can be offset by optimizing computing efficiency or network configuration, such as halving the LPDDR5 capacity per rack while still meeting overall rack requirements. To suit different enterprise application scenarios, Nvidia promoted its open-source model Nemotron, a tool that helps clients customize their own full-stack AI solutions.
