Oppenheimer analyst Rick Schafer pointed out in a note to clients that Nvidia's revenue guidance for the fourth fiscal quarter (as of January) and the first fiscal quarter of 2026 (as of April) has a "typical" upside potential of $2 billion to $3 billion, mainly benefiting from the strong demand for the GB300 Ultra product line. Wall Street's current revenue estimates for the above two quarters are $65.6 billion and $71.6 billion, respectively.
Schafer said that the capital expenditure of cloud service providers continues to rise, and it is expected that the capital expenditure of global cloud vendors will reach 650 billion USD in 2026, significantly higher than the level of more than 400 billion USD in 2025. At the same time, the scale of cutting-edge large models (LLMs) is still growing at a rate of about 10 times per year, and the demand for inference tokens is also growing at a rate of more than 5 times, which further pushes up the demand for high-performance AI computing power.
At the product level, Schafer noted that Nvidia's rack-level solution NVL72 remains an industry leader in AI performance per unit power consumption, while the new-generation Vera Rubin (VR200) platform is progressing as planned and is expected to achieve a mass production climb in the third quarter of fiscal 2026, followed by a higher-end VR300 Ultra expected to launch in the early third quarter of fiscal 2027.
Schafer further estimates that the average selling price of the Vera Rubin platform is expected to be 40% to 50% higher than the GB300. For reference, the GB300 costs about $3.5 million in a single set. Based on this, the Vera Rubin series of products is expected to bring about $8 billion in new revenue to Nvidia in the future.
In addition, with the reintegration of the Chinese market into the serviceable scope, the potential reachable market size may reach as high as US$50 billion, which may further push up Nvidia's overall serviceable market size (TAM), which is currently estimated to be about US$4 trillion. In the long run, Nvidia remains "the most versatile AI platform and the most winning player," Schafer said.