Nvidia just told Wall Street that the AI spending wave isn’t slowing down. It’s accelerating.
The chipmaker reported fiscal Q2 2027 revenue of $96.2 billion on August 26, a 106% increase from the same period a year ago. More importantly, Nvidia issued its first-ever full-year revenue guidance, projecting roughly 70% growth for fiscal 2028. Analysts had been expecting around 44%.
The stock jumped more than 4% in after-hours trading.
The data center engine
Nvidia’s data center segment pulled in $89 billion during the quarter, up 117% year-over-year and representing roughly 92% of total company sales.
CEO Jensen Huang and CFO Colette Kress pointed to diversification within that demand as a key signal. Non-hyperscale customers, meaning companies outside the usual suspects like Microsoft, Google, and Amazon, now account for about half of Nvidia’s business. That cohort is growing at nearly 100% year-over-year.
What $673 billion in revenue would mean
If Nvidia hits that 70% growth target for fiscal 2028, implied revenue would land somewhere near $673 billion, which would make Nvidia the second-largest technology company in the US by revenue, behind only Amazon.
Consider where Nvidia was just a few years ago. The company’s annual revenue was in the $27 billion range for fiscal 2023. Growing from that to potentially $673 billion in five fiscal years would represent one of the most dramatic revenue scaling stories in corporate history, in any industry.
Margins and supply constraints
Nvidia expects gross margins to stabilize around 71% to 72% in the coming quarters. The compression stems from memory costs: Nvidia’s latest-generation AI chips require high-bandwidth memory (HBM), and supply of those components remains tight.
Huang and Kress acknowledged that supply constraints, particularly in memory, remain a key challenge. This is a company that could likely sell more chips than it can currently build.
A 71% to 72% gross margin would still be remarkable for a hardware company operating at this scale. For context, Apple’s gross margins hover around 45%. Intel’s have been compressed well below 50%.
What to watch from here
The non-hyperscale growth is arguably the more important signal for the medium term. When enterprise customers outside Big Tech start spending at this pace, it typically indicates that AI workloads are moving from experimental to production.
The memory supply constraint could also become a more significant factor if HBM production doesn’t scale as quickly as chip demand. Samsung, SK Hynix, and Micron are all expanding HBM capacity.
Nvidia’s guidance beat of 70% versus the 44% consensus represents the widest positive gap between company outlook and analyst expectations in recent memory for a company of this size.