Nvidia's Q2 Revenue Exceeds Expectations, Data Center Business Booms, Blackwell GPU Set for Mass Production

Nvidia's Q2 Revenue Exceeds Expectations, Data Center Business Booms, Blackwell GPU Set for Mass ProductionOn August 29, Nvidia released its second-quarter fiscal 2025 earnings report for the period ending July 28. The report revealed revenue of $30 billion, a 122% year-over-year increase, surpassing analyst expectations of $28

Nvidia's Q2 Revenue Exceeds Expectations, Data Center Business Booms, Blackwell GPU Set for Mass Production

On August 29, Nvidia released its second-quarter fiscal 2025 earnings report for the period ending July 28. The report revealed revenue of $30 billion, a 122% year-over-year increase, surpassing analyst expectations of $28.7 billion. Net income amounted to $16.6 billion, representing a 168% year-over-year growth. Diluted earnings per share reached $0.67, marking a 168% year-over-year increase, topping analyst estimates of $0.64.

Following the release, Nvidia CEO Jensen Huang, CFO Colette Kress, and other executives participated in the earnings call, discussing the report and answering analyst questions.

Kress described the quarter as another record-breaking period, with Nvidia revenue reaching $30 billion, reflecting a 15% sequential and 122% year-over-year surge, significantly exceeding their initial forecast of $28.7 billion. Key highlights included robust growth in the data center business, which generated $26.3 billion in revenue, not only a new record but also a 16% sequential and 154% year-over-year increase. This growth was largely fueled by strong demand for Nvidia's Hopper GPU computing and networking platforms. Computing revenue more than tripled, while networking revenue doubled. Cloud service providers contributed roughly 45% to data center revenue, with the remaining 50% coming from consumer internet and enterprise sectors.

The robust demand for Hopper GPUs continues to drive business growth, while customers are actively preparing for the adoption of the next-generation Blackwell products. The next generation of models will necessitate 10 to 20 times more computing power to process larger datasets, a trend projected to persist. Nvidia's demand is diverse and widespread, encompassing cutting-edge model developers, consumer internet services, and tens of thousands of companies and startups. These clients are leveraging Nvidia's technology and ecosystem to develop generative AI applications across various fields, including consumer, advertising, education, enterprise, healthcare, and robotics.

Nvidia's H200 platform entered mass production in the second quarter, supplying major cloud service providers, consumer internet companies, and enterprises. During the quarter, Nvidia's data center business saw sequential growth in the Chinese market, emerging as a significant contributor. However, despite this growth, the revenue share from China remains below pre-export restriction levels. Nvidia anticipates intense competition in the Chinese market.

Demand for Blackwell GPUs remains robust, prompting Nvidia to refine its production processes for enhanced output. The production expansion plan will commence in the fourth quarter of fiscal 2025 and continue into fiscal 2026. Nvidia expects significant Blackwell shipments in the fourth quarter, generating billions in revenue. While supply and availability have improved with increasing shipments in the latter half of fiscal 2025, demand still outpaces supply, a situation the company expects to persist into next year.

As countries increasingly recognize the strategic importance of AI expertise and infrastructure for societal progress and industrial upgrades, the opportunity for sovereign AI continues to expand. Nvidia anticipates significant revenue growth, amounting to billions of dollars, in the sovereign AI domain this year. Concurrently, an enterprise AI wave is gaining momentum, with numerous companies actively driving continuous sequential revenue growth. Nvidia is collaborating with numerous Fortune 100 companies to advance cross-industry and cross-regional AI innovation projects.

Among the numerous growth-driving sectors, the automotive industry stands out as a key engine for Nvidia's quarterly growth. Automotive manufacturers are accelerating their deployment of autonomous vehicle technologies, deeply integrating AI into data centers. This trend not only contributes billions in revenue but also signifies further growth potential as the computational demands of next-generation autonomous driving models surge.

In the gaming and AIPC domains, the gaming segment displayed strong market vitality, generating $2.88 billion in revenue during the second quarter, representing a 9% sequential and 16% year-over-year increase. Both gaming console and laptop markets experienced sequential growth, with sustained strong and stable demand. Channel inventory remains at healthy levels. In professional visualization, revenue reached $454 million, signifying a 6% sequential and 20% year-over-year increase. This growth is primarily driven by the deep integration of AI and graphics technologies, particularly in the widespread adoption of model optimization and innovative workloads like Omniverse. Additionally, Nvidia generated $346 million in revenue from automotive and robotics, reflecting a 5% sequential and 37% year-over-year increase. This growth is attributed to strong demand for autonomous driving platforms from new customers and the widespread adoption of AI-powered cockpit solutions.

During the Q&A session, Huang stated that Nvidia has completed mask improvements for Blackwell, eliminating the need for further design adjustments from a functional perspective. Currently, Nvidia is fully committed to functional sample testing of Grace Blackwell in various system configurations.

Huang further emphasized the accelerating diversification and rapid growth of generative AI, stating that investing in Nvidia infrastructure offers immediate and significant returns, representing the most optimal ROI for infrastructure computing investments.

Regarding Blackwell's production ramp-up, Huang mentioned that Nvidia provides diverse configuration options, including the classic version (based on HGX form factor) and the liquid-cooled version (Grace Blackwell). The adoption of liquid cooling technology significantly enhances data center AI throughput, enabling the deployment of more computing power within the same space, while simultaneously reducing costs.

Kress expressed Nvidia's confidence in continued Hopper growth during the second half of the year, projecting significant performance improvements starting from the third quarter and continuing into the fourth quarter. Additionally, the ramp-up of Blackwell production in the fourth quarter will further boost Nvidia's revenue growth.