NVIDIA Earnings: The AI Supercycle Is Still Accelerating

‌$英伟达(NVDA)$  ‌ once again proved that AI infrastructure demand is far from slowing down.

Key numbers:

• Revenue: $96.22B vs ~$92.17B (beat)

• Adjusted EPS: $2.22 vs $2.09-$2.10 (beat)

• Data Center Revenue: $89.0B vs ~$86.3B (beat)

• Adjusted Gross Margin: 75% (in line)

The most important part:

Hyperscaler revenue reached:

$48.71B

vs $43.05B last quarter

Despite the market’s concerns about custom ASIC competition, major cloud providers are still aggressively investing in AI infrastructure, and NVIDIA continues to capture the majority of this spending.

Forward guidance:

• Revenue: $108B vs ~$104.2B (beat)

• Gross Margin: 74% (slightly softer)

The only minor weakness was the expected gross margin decline.

But the revenue ramp remains extraordinary:

$68.1B

→ $81.6B

→ $96.2B

→ $108B next quarter guidance

And this guidance assumes zero China Data Center compute revenue.

One of the most interesting details:

NVIDIA’s purchase commitments increased from $119B last quarter to $279B, mainly related to future memory procurement.

Translation:

The AI boom is not only a GPU story.

It is creating a massive demand cycle across the entire infrastructure stack:

GPU → HBM → Memory → Networking → CPO → Power Infrastructure

My view:

At a $5T+ market cap, the biggest opportunities may no longer come from simply finding mispricing in NVIDIA itself.

The bigger opportunity is understanding how NVIDIA’s architecture decisions, supply chain requirements, and capacity expansion impact the rest of the AI ecosystem.

The winners of the AI era will not only be GPU companies.