NVIDIA Revenue Rises 106%: Growth May Be Its Own Biggest Enemy

An impressive earnings report

NVIDIA's second-quarter 2026 results were exceptionally strong again: revenue reached $96.2 billion, up 106% from a year earlier. Data Center revenue was $89.0 billion, accounting for 92% of the total. Third-quarter revenue guidance of $108.0 billion also continued to exceed market expectations.

After the earnings release, NVIDIA shares initially fell before rebounding, ending the session up 8.74%.

Investors were caught between two concerns: that the rally had borrowed too much from future growth and valuations were inflated, and that they could not afford to miss the AI trade.

Enemy one: surging costs

NVIDIA's CFO identified a central problem: AI construction itself was largely responsible for the shortage of memory.

That shortage also raises NVIDIA's own costs.

Third-quarter gross-margin guidance was 74%, with the fourth quarter expected to fall to 71%-72%. Rising costs continued to squeeze profits. Even though the company expected price increases to restore margins in the first quarter of next year, near-term margin pressure was already in place.

Enemy two: global inflation

The rapid expansion of AI infrastructure is not merely increasing costs within the industry. It is also creating inflationary pressure more broadly.

Large-scale data-centre construction and growing demand for computing power and electricity raise prices for memory, electricity and computing resources, eventually feeding into prices across the economy.

The US core PCE inflation reading of 3.3% illustrates this mechanism in the author's analysis; it is an inference about the mechanism, not an official attribution.

This is a structural paradox for NVIDIA. With 92% of its revenue coming from data centres, it is AI's biggest beneficiary and a major driver of this round of AI-related inflation. Faster growth brings greater inflationary pressure.

Enemy three: interest rates and emerging competition

The inflation persistence created by AI constrains the Fed's room to cut rates, keeping interest rates high for longer.

NVIDIA trades at a high valuation and is highly sensitive to interest rates. In effect, the AI growth it helps create delays rate cuts and puts pressure on its own valuation.

Meanwhile, another development is emerging within the industry. OpenAI says its internally developed Jalapeño chip has outperformed GB200/GB300 in its own tests. It is not yet in mass production, and those claims have not been independently verified. It is unlikely to challenge NVIDIA's dominance in the short term, but the move towards internally designed alternatives has begun.

AI's gains now need to be weighed against its risks: three things to watch

1. Gross margins: is 71%-72% a temporary floor or the start of a decline?

Whether price increases can restore margins in the first quarter of next year is the key near-term variable.

2. Memory supply and demand: when the shortage eases will directly determine when cost pressure peaks and margins stabilise.

3. Cloud providers' ordering patterns: major buyers are still building inventories aggressively. If leading customers cut orders, NVIDIA's high-growth narrative could reverse.

AI's extraordinary popularity is not a free benefit; it comes with a bill paid over time

Margins are under pressure in the short term, while inflation and interest rates create medium-term constraints. In the quarters ahead, growth alone will no longer decide the outcome. The ability to absorb risk will matter, and the deciding factors will soon become clear.