Chipmaker Nvidia posted record quarterly revenue of $68 billion, marking a 73% surge from a year earlier as AI demand accelerated across global markets.

The world’s most valuable chip company generated $62 billion of that total from its data center division, underscoring the dominance of AI-driven compute.

Within the segment, $51 billion came from GPU compute while $11 billion was attributed to networking technologies such as NVLink.

Full-year revenue reached an extraordinary $215 billion, reinforcing Nvidia’s commanding position in the AI hardware ecosystem.

CEO Jensen Huang told investors that global demand for AI “tokens” remains overwhelming, with even older cloud GPUs fully utilized and pricing climbing.

The company confirmed it recorded no revenue from chip exports to China during the quarter, reflecting ongoing regulatory restrictions.

CFO Colette Kress noted that limited shipments of H200 chips received U.S. approval but have not yet generated revenue, while Chinese rivals such as Moore Threads continue to advance.

Huang also addressed Nvidia’s reported $30 billion pending investment in OpenAI, stating that discussions are progressing toward a potential partnership.

The company maintains collaborations with AI leaders including Meta, Anthropic, and Elon Musk’s xAI as it expands its ecosystem footprint.

Despite optimism, filings with the U.S. Securities and Exchange Commission cautioned there is no assurance the OpenAI investment will ultimately materialize.

Huang described the industry as reaching an inflection point, arguing that in the AI economy “compute is revenue,” and that profitable token generation will drive sustainable long-term growth.