Overview
- On August 26 Nvidia reported Q2 revenue of about $96.2 billion with Data Center sales near $89 billion and issued its first-ever year‑ahead forecast calling for roughly 70% revenue growth for fiscal 2028 that management says is limited by supply rather than demand.
- Nvidia says the Vera Rubin GPU is now in production and that orders from major cloud providers are powering the surge, with management estimating hyperscaler capex near $1.3 trillion in 2027 and projecting multi‑trillion dollar annual AI infrastructure spending later this decade.
- Memory and advanced packaging shortages have pushed component prices higher and pressured gross margins, and Nvidia disclosed sharply larger multiyear supply and financing commitments that increase its exposure to suppliers and a few large customers.
- The strength is broadening across the ecosystem: Broadcom reported AI chip revenue that more than tripled, and OEMs including Dell and HPE showed record AI server sales and backlogs, signaling the industry is supply‑constrained rather than demand‑starved.
- Key near‑term risks for investors are concentrated customer exposure, rising memory costs that can compress profitability, geopolitical limits such as China exclusions in guidance, and growing scrutiny of Nvidia’s large financing arrangements and contingent liabilities.