NVIDIA $NVDAB has still submitted a completely unreasonable set of results: FY2027 Q2 revenue was $96.221 billion, up +106% year over year; data center revenue was $89.0 billion, up +117% YoY; GAAP net profit was $59.688 billion, up +126% YoY. Growth on this scale is indeed wonderful for AI. In my view, it should be analyzed like the financial report of a mature industry—looking at revenue, gross margin, accounts receivable, capital returns, and next quarter’s guidance.
For example, in Q2, net accounts receivable were $63.059 billion, up from $38.466 billion at the beginning of the period. Inventories were $31.575 billion, up from $21.403 billion at the beginning of the period. You see, when revenue grows extremely fast, working capital expands too. If you only focus on revenue and the stock price, you’ll miss the capital tied up by business expansion.
For instance, in Q2 the company returned about $26 billion to shareholders in the form of buybacks and dividends, and the remaining share repurchase authorization at quarter-end was about $99.0 billion. A qualified earnings report should look like this—no matter whether cash is put into R&D, acquisitions, buybacks, dividends, debt repayment, or cash hoarding, ordinary investors need the company to clearly communicate capital allocation.
Finally, strong earnings like those from a leading company such as Nvidia prove its own execution and order capabilities, but they can’t be used to characterize the entire AI industry chain. Upstream, data centers, power, cloud services, and the application layer all have very different cash conversion cycles and competitive dynamics. After all, anyone can shout slogans. And just like Nvidia’s next-quarter revenue guidance of $108.0 billion ±2%, whether that number is reliable ultimately still comes down to DYOR#英伟达营收超预期股价涨4%
For example, in Q2, net accounts receivable were $63.059 billion, up from $38.466 billion at the beginning of the period. Inventories were $31.575 billion, up from $21.403 billion at the beginning of the period. You see, when revenue grows extremely fast, working capital expands too. If you only focus on revenue and the stock price, you’ll miss the capital tied up by business expansion.
For instance, in Q2 the company returned about $26 billion to shareholders in the form of buybacks and dividends, and the remaining share repurchase authorization at quarter-end was about $99.0 billion. A qualified earnings report should look like this—no matter whether cash is put into R&D, acquisitions, buybacks, dividends, debt repayment, or cash hoarding, ordinary investors need the company to clearly communicate capital allocation.
Finally, strong earnings like those from a leading company such as Nvidia prove its own execution and order capabilities, but they can’t be used to characterize the entire AI industry chain. Upstream, data centers, power, cloud services, and the application layer all have very different cash conversion cycles and competitive dynamics. After all, anyone can shout slogans. And just like Nvidia’s next-quarter revenue guidance of $108.0 billion ±2%, whether that number is reliable ultimately still comes down to DYOR#英伟达营收超预期股价涨4%