NVIDIA has cemented its position as the backbone of global artificial intelligence development, reporting record quarterly revenue of $46.7 billion, up 56% year-on-year. CEO Jensen Huang said the company’s Blackwell and Rubin AI factory platforms will be central to the “$3 trillion to $4 trillion global AI factory build out through the end of the decade.” The firm is now manufacturing 1,000 AI racks each week and expects revenue of $54 billion next quarter, excluding sales to China.
China continues to represent both a major opportunity and a growing competitive front. NVIDIA CFO Colette Kress disclosed shipments of $2–5 billion worth of H20 GPUs to the country. Huang estimated that China’s AI chip market could hit $50 billion in 2024 alone, growing at 50% annually. At the same time, domestic players are intensifying efforts to reduce reliance on foreign technology. Huawei’s Ascend 910C and 920 chips are emerging as strong contenders, while Alibaba, Baidu, and Cambricon Technologies are also scaling their GPU initiatives—Cambricon recently reported profits of Rmb1 billion ($140 million).
Meanwhile, demand from governments worldwide is accelerating NVIDIA’s growth. As sovereign AI ambitions expand, countries are pouring billions into localised AI infrastructure. The EU has pledged €20 billion for 20 AI factories, the UK has deployed its most powerful supercomputer using NVIDIA hardware, and the Middle East is investing heavily in AI data centres. Kress noted that sovereign AI revenue will exceed $20 billion this year, doubling 2023’s figure. She added: “Nations are treating AI as a matter of national security, and NVIDIA is cashing in.”
Looking to the future, Huang said the next evolution of AI will be reasoning systems capable of planning, tool use, and independent research—far beyond today’s chatbots. Such systems could require 100x–1000x more compute, placing NVIDIA’s Blackwell NVLink 72 rack-scale platform at the heart of the transition. “In a world of power-limited data centres, perf per watt drives directly to revenues,” Huang emphasized. While rivals pursue custom ASICs, he argued NVIDIA’s combination of GPUs, CPUs, networking, and CUDA software creates a full-stack ecosystem unmatched in accelerated computing
