The dust settles on Nvidia: Morning Brief

Brian Sozzi · Executive Editor

The dust has settled on the latest Nvidia (NVDA) earnings week frenzy.

And I think it's important to take stock of where things stand for the world's most important stock (sorry, Apple (AAPL)). Why? Because you should be thinking about whether the pullback in Nvidia is a buying opportunity or the start of a deeper sell-off as expectations are reset.

We know Nvidia's margins in the first half of the year will be below their usual robust levels as Blackwell AI chips ramp up. I would argue the Street knew this ahead of the results, so they got flustered over nothing.

On Nvidia's earnings call, execs sought to push back on the bears, who have put forth a narrative that there will be a digestion period for AI investments by hyperscalers such as Amazon (AMZN) and that Nvidia's margins may have peaked.

"Once our Blackwell fully rounds, we can improve our cost and our gross margin," Nvidia CFO Colette Kress said. "So, we expect to probably be in the mid-70s later this year."

We also know that, fundamentally, Nvidia's business is strong and likely to stay strong.

Fourth quarter revenue rose 12% sequentially and 78% from the prior year. Data center sales more than doubled from the prior year. Earnings handily beat analyst estimates.

"We're going to have to continue to scale as demand is quite high, and customers are anxious and impatient to get their Blackwell systems," Nvidia founder and CEO Jensen Huang said.

Nowhere in the company's 2025 guidance or commentary from Huang did I sense that AMD (AMD) is taking Nvidia's market share; ditto custom chips from Amazon. I heard no hint that hyperscalers are sending AI chips back to Nvidia or have stopped fawning over Jensen to get more of these chips at any cost.

Put together, I would argue what we heard from Nvidia in terms of demand and margins was all well known going into the results. So, the sell-off could prove to be an overreaction, a function of investors aiming to model out mixed first quarter guidance for the next two years.

But there are a couple of things we don't yet know about Nvidia that warrant greater attention. These play into the long-term bull thesis.

For starters, there's Huang's point about DeepSeek's R1 requiring 100x more compute resources compared to pre-training models due to inference time scaling. Look, most of us have no clue what this even means. But the casual observer could read it as the market has it strong on DeepSeek, and there could be a lot of upside to Nvidia estimates as DeepSeek and other reasoning models gain hold.

And the second thing we don't know is the long-term impact of what Huang will show off at Nvidia's GTC event on March 17.

"We're going to provide a big, huge step-up [in performance]," Huang said. "And so, come to GTC, and I'll talk to you about Blackwell Ultra, Vera Rubin, and then show you the one click after that. Really exciting new products to come at GTC."

To me, these new chips could blow Blackwell's performance away and reinforce Nvidia's leading position in the space.

The last thing that is unclear is how nation-states' buildout of AI infrastructure will drive demand for Nvidia chips. A16z general partner Anjney Midha suggests the Street may be undervaluing the opportunity.

I am not some crazy Nvidia bull. But I am a practitioner of commonsense when studying stocks, companies, and leaders. And when it comes to Nvidia, it's silly to think what we heard from the company is going to send the stock to end 2025 lower.