Key Takeaways:
Nvidia reported fiscal Q2 revenue of $96.2 billion, up 106 percent year over year, beating consensus by $4 billion as AI demand accelerated.
"AI has reached its inflection point. It's doing useful work. Its tokens are productive and profitable. Now, compute is revenue," Jensen Huang, founder and CEO of Nvidia, said in a statement. "And demand is accelerating."
Data center revenue reached $89 billion, beating estimates of $85.7 billion. Hyperscale customers contributed $48.7 billion, up 13 percent sequentially, while the ACIE segment — AI clouds, industrial and enterprise — added $40.3 billion, up 138 percent year over year. Non-GAAP EPS was $2.22, above the $2.06-$2.09 consensus range.
Shares rose more than 5 percent in after-hours trading as management guided third-quarter revenue to $108 billion, plus or minus 2 percent, and projected approximately 70 percent growth in fiscal 2028 — a pace CFO Colette Kress said is supply-constrained.
Kress said Nvidia's Vera Rubin platform began production shipments in August and is expected to account for about 20 percent of third-quarter data center revenue, making it the fastest product ramp in company history. AWS plans to deploy an additional two million Nvidia GPUs through the second quarter of fiscal 2029, including Vera CPUs integrated with Rubin products.
Nvidia's networking revenue hit a record, up 18 percent sequentially, with Spectrum-X Ethernet revenue growing 2.6 times year over year. NeoCloud partners using Nvidia's DSX reference designs are expected to exit the year with eight gigawatts of installed capacity, up from roughly three gigawatts at the end of 2025.
Gross margin held at 75 percent in the quarter but is expected to decline to 71-72 percent in the fourth quarter because of rising memory costs before stabilizing at 72-73 percent in fiscal 2028. Inventory rose to $32 billion as Nvidia prepared for the Vera Rubin launch.
The company returned a record $26 billion to shareholders, including $20 billion in buybacks and $6 billion in dividends, with $99 billion remaining in buyback authorization. Nvidia shipped less than 1 percent of data center revenue in Hopper 200 products to China-based customers during the quarter, and its forward outlook includes no China data center compute revenue because of geopolitical uncertainty.
Huang said Nvidia's revenue opportunity per gigawatt of data center capacity has grown from roughly $18 billion in the Hopper generation to $25 billion with Blackwell and $40 billion with Vera Rubin, including CPUs, GPUs, networking, systems and software. The company has invested nearly $50 billion in frontier AI laboratories and partnered with Apollo Global Management, BlackRock, Blackstone, Brookfield Asset Management, Goldman Sachs and KKR on financing platforms expected to raise more than $500 billion in third-party capital.
The guidance raise implies AI demand will stay above supply through fiscal 2028. Investors will watch the third-quarter earnings call for updated gross margin trends and Vera Rubin ramp details.
This article is for informational purposes only and does not constitute investment advice.
