英伟达市值突破万亿美元

2023 年 5 月，英伟达的市值在财报发布后盘中突破 1 万亿美元，成为历史上第一家达到这个高度的芯片公司。对很多普通用户来说，这个名字还停留在「显卡」的认知里；但对整个 AI 行业来说，这一刻早已注定——当全世界的公司都在抢购 GPU 训练大模型时，那个造 GPU 的厂商，正在变成这场淘金热里唯一的「卖铲人」。

英伟达的爆发不是一夜之间的事。数据中心业务多年的积累、CUDA 生态的绑定、以及每一代芯片在算力上的领先，都为这一刻铺垫了许久。但真正的引爆点，是 2022 年底 ChatGPT 带来的生成式 AI 浪潮。训练和推理大模型需要海量 GPU，需求远超过供给，英伟达的芯片成了供不应求的战略物资。

万亿市值意味着资本市场完成了一次重新定价：AI 不再只是软件公司的故事，而是芯片、数据中心、能源与电力串起来的全链条产业。GPU 从「游戏配件」被重新定义为「AI 时代的基础设施」，就像石油之于工业时代。这种认知转变深刻影响了此后两年的产业叙事——算力供应链成为地缘与资本竞争的焦点，各国都把芯片视为战略资源。

当然，万亿市值也带来了新的议题。算力集中引发的依赖与安全担忧、GPU 短缺对中小开发者的挤出效应、以及「卖铲人」与「淘金者」之间的利益结构，都成为行业持续讨论的话题。英伟达的崛起既证明了 AI 算力的价值，也让「算力鸿沟」成为新的不平等来源。

回看英伟达破万亿的那一刻，它像一个分水岭：在此之前，AI 谈论的是算法与模型；在此之后，AI 开始谈论芯片、电力与供应链。当 2026 年的数据中心还在为 GPU 和电力激烈竞争时，2023 年 5 月那个万亿市值，正是这一切的开端。

In May 2023 NVIDIA's market value broke $1 trillion intraday after an earnings report, making it the first chip company ever to reach that height. To many ordinary users the name still meant "graphics cards"; to the AI industry, though, this moment was a long time coming—when the whole world is rushing to buy GPUs to train large models, the maker of those GPUs becomes the only "shovel seller" in the gold rush.

NVIDIA's eruption was not overnight. Years of accumulation in the data-center business, the lock-in of the CUDA ecosystem, and a lead in compute in each new generation all paved the way. But the true trigger was the generative-AI wave ignited by ChatGPT in late 2022. Training and serving large models demands GPUs at a scale supply could not meet, and NVIDIA's chips became a scarce strategic commodity.

The trillion-dollar valuation meant the capital market had finished repricing a shift: AI was no longer just a software story, but a full-chain industry of chips, data centers, energy, and power. GPUs were redefined from "gaming parts" to "the infrastructure of the AI era," like oil for the industrial age. That cognitive shift shaped the next two years of industry narrative—the compute supply chain became a focus of geopolitical and capital competition, and every country treated chips as a strategic resource.

Of course, the trillion-dollar mark also opened new questions. Concerns about concentration and dependence, the squeeze on smaller developers from GPU shortages, and the interest structure between the "shovel seller" and the "miners" all became ongoing industry topics. NVIDIA's rise proved the value of AI compute while also making the "compute divide" a new source of inequality.

Looking back at the moment NVIDIA passed a trillion dollars, it stands as a watershed: before it, AI talked about algorithms and models; after it, AI started talking about chips, power, and supply chains. When data centers in 2026 are still fighting over GPUs and electricity, that trillion-dollar mark in May 2023 is where it all began.