The hardware infrastructure: Nvidia demonstrates the real profitability of AI
Nvidia has released its report for the second quarter of its fiscal year 2027, significantly surpassing even the most optimistic forecasts from Wall Street and sending a strong signal regarding the profitability of the artificial intelligence sector.
A quarter of financial records
During the period ending on July 26, 2026, the technology company achieved revenues of $96.2 billion, a 106% year-on-year increase and an 18% rise compared to the previous quarter. This figure exceeded the market analysts' forecasts, who expected revenues around $92.17 billion.
The growth was reflected uniformly across all its key profitability indicators:
- Earnings per share: It stood at $2.22 adjusted, surpassing the $2.10 projected by the market and marking a year-on-year growth of 128%.
- Net income: Reached $59.688 billion, equivalent to a 126% annual increase.
- Adjusted gross margin: Rose to 75.0%, compared to the 72.4% recorded in the same period the previous year, while operating revenues touched $63.734 billion.
Data centers and the memory bottleneck
The main driver of the business remains the data center division, whose revenues reached $89 billion, registering a remarkable year-on-year increase of 117%. "AI has reached its turning point. It is doing useful work. Its tokens are productive and profitable. Now, computing is revenue," stated Jensen Huang, founder and CEO of Nvidia.
However, the operational success exposes the physical pressure on the supply chain. The report reveals that Nvidia's supply commitments doubled, rising from $119 billion the previous quarter to $279 billion in this period, an increase primarily associated with the acquisition of memory chips due to market shortages.
Despite these historic results, the company's shares experienced an initial drop of 1.59% at the close of the trading session, settling at $209.66, although they rebounded by 4.08% in after-hours trading to quote at $218.21. For the third fiscal quarter, the company projects revenues of $108 billion.
For next+ Nvidia's financial report dispels doubts about whether artificial intelligence is a speculative bubble or a real business. When the basic hardware to run these models generates margins of 75.0% and growth forecasts nearing $108 billion quarterly, the discussion shifts from the software's potential to the physical capacity of the supply chain. For decision-makers in Latin America, the true indicator to monitor is no longer just Big Tech's spending on AI, but the shortage of specialized memory components, a logistical bottleneck that will define the speed of local infrastructure deployment during the second half of the decade.