Nvidia's fiscal 2027 second-quarter revenue reached $96.2 billion, more than double a year ago, but the company guided for a lower gross margin ahead. Chief Financial Officer Colette Kress pointed to rising memory costs as the cause, a shift that puts pricing power in the hands of memory suppliers.
Nvidia's total revenue reached $96.2 billion in its fiscal 2027 second quarter, more than double the $46.7 billion posted a year earlier and up 18% from the prior quarter. The data center segment alone generated $89 billion, more than the company's entire revenue one year ago. Within that segment, hyperscale customers contributed $48.7 billion, an increase of 102% year over year.
Gross margin came in at 75%, up 2.6 points from a year ago, while operating income rose 19% sequentially and 124% year over year to $63.7 billion. Net income reached $59.7 billion, or $2.46 per share. For the third quarter, Nvidia expects revenue of $108 billion, plus or minus 2%, and the guidance assumes no compute sales to China.
Nvidia flags a margin step-down ahead
Nvidia's outlook carries a softer signal beneath the growth. GAAP and non-GAAP gross margin are both expected to be 74%, plus or minus 50 basis points, a full point below the margin the company just delivered. Management said gross margin is expected to bottom between 71% and 72% by the fourth quarter before recovering in fiscal 2028.
Kress was direct about the cause: rising memory costs, not weaker demand or fading pricing power. She said commitments for critical components increased from $119 billion last quarter to $279 billion, adding that the jump was "primarily related to the procurement of memory."
High-bandwidth memory turns into a co-equal cost
High-bandwidth memory has moved from a commodity add-on to a co-equal part of the accelerator package, driving up what Nvidia must pay to secure supply. The article's author sees Nvidia's margin cut as a positive signal for memory suppliers, arguing that a customer as large as Nvidia accepting a lower margin outlook over memory costs confirms the bottleneck is moving into DRAM and HBM production. In the author's view, the same hyperscaler spending cycle driving Nvidia's growth may now be flowing into higher realized prices for the memory suppliers on the other side of its purchase orders.
Source: The Motley Fool
Trading involves risk.
Risk Disclaimer
