Hey there, bargain hunter.

Nvidia just reported the most dominant quarterly earnings print in semiconductor history, and the stock slipped 1% in after-hours. Let that sit for a second.

That’s not a red flag. That’s what happens when expectations are so stratospheric that beating them starts to feel routine. The bar doesn’t just move – it levitates.

What the Numbers Actually Say

Revenue came in at $81.6 billion for fiscal Q1 – up 85% year-over-year, up 20% sequentially, and comfortably ahead of the $78.8 billion Wall Street was penciling in. Adjusted EPS hit $1.87, clearing the consensus range of $1.76–$1.78. Gross margin held at 75%, which is not a given at this revenue scale. These are not soft beats. These are decisive ones.

The data center segment – the engine behind all of this – grew 92% year-over-year to $75.2 billion. Management flagged accelerating demand for Blackwell GPU architecture specifically. When your core segment nearly doubles and you’re still supply-constrained, the story isn’t slowing down.

The Q2 guide is where things get genuinely interesting. Nvidia projected $91 billion at the midpoint – well ahead of the $86.1 billion analysts expected. The detail that matters most: that guide assumes zero data center compute revenue from China. Zero. They’re growing through the export restriction, not around it.

Capital Allocation – This Is Not a Small Company Acting Small

An additional $80 billion share repurchase authorization. A dividend hike from $0.01 per share to $0.25 per quarter. In Q1 alone, Nvidia returned roughly $20 billion to shareholders via buybacks and dividends. That’s not a growth company hedging – that’s a cash machine running at full tilt.

Slight tangent, but worth noting: the dividend increase is 25x in one move. That kind of signal from management isn’t accidental. They’re telling you the cash flow trajectory is durable, not cyclical.

The Framework Shift Nobody Is Talking About

Nvidia is restructuring how it reports its business. Data Center splits into Hyperscale (public clouds, large internet platforms) and ACIE – AI Clouds, Industrial, and Enterprise. Edge Computing becomes a single umbrella covering PCs, gaming consoles, robotics, and automotive AI. This isn’t cosmetic. When a company rewrites its reporting segments, it’s usually because the old buckets no longer reflect where revenue is actually going.

Is It Cheap?

Depends on which lens you use. Trailing P/E sits around 45x – about 35% above the broader tech sector average of ~33x. Expensive on its face. But trailing earnings are a rearview mirror after a quarter like this one.

The forward picture flips the narrative. The forward P/E is approximately 27x – meaningfully below the semiconductor industry median of ~36x. The most dominant chip company on earth is trading at a discount to its own industry on a forward basis. That’s the number that matters after a guide like $91 billion.

One more data point worth holding: NVDA’s current trailing P/E is roughly 18% below its own 10-year historical average of ~54x. Investors have paid far more for this stock in prior cycles when growth was a fraction of what it is today.

Cheap? Not in the traditional sense. Cheaper than it looks? The data says yes. The debate isn’t whether Nvidia is expensive – it’s whether the growth runway justifies the price. After tonight, that runway just got longer.