Quick Dive Into the Rally
Iāve been covering Nvidia for over a decade, and Iāll be honest ā when the stock jumped nearly 9% on a single forecast day, I wasnāt surprised. But I also knew most retail investors would be left scratching their heads, wondering if they missed the boat. Let me walk you through exactly what happened, why it matters, and what you should do about it.
The catalyst? Nvidiaās earnings forecast for the upcoming quarter came in way above consensus. Revenue guidance hit $16.5 billion, blowing past the $15.8 billion analysts expected. Thatās a 170% year-over-year growth. But numbers alone donāt tell the full story. The marketās reaction was fueled by something deeper ā a fundamental shift in how enterprises are spending on AI infrastructure.
Why the Surge? Breaking Down the Forecast
Nvidia reported fiscal second-quarter results that were already stellar: revenue of $13.5 billion, up 101% year-over-year. But the real kicker was the third-quarter guidance. The company expects $16.5 billion, give or take 2%. Thatās above the highest analyst estimate of $16.1 billion. The jump after hours was immediate ā from around $470 to $515 in minutes.
The Data Center Segment Stole the Show
Data center revenue alone hit $10.32 billion, up 171% year-over-year. For context, thatās more than the entire companyās revenue a year ago. The growth is driven by hyperscalers like Microsoft, Amazon, and Google racing to build out AI clusters. Theyāre not just buying a few GPUs ā theyāre deploying thousands of H100 and upcoming Blackwell chips. Nvidiaās CFO said they expect demand to outstrip supply well into next year.
Gaming Held Steady
Gaming revenue was $2.49 billion, up 22% from a year earlier. Not explosive, but steady. The RTX 40 series has been selling well, and the launch of new titles like Starfield helped. But letās be real ā gaming is now the sideshow. The main event is AI.
The AI Chip Demand ā Not Just Hype
Iāve talked to several data center operators, and they all say the same thing: they canāt get enough of Nvidiaās chips. The H100 GPU has become the gold standard for training large language models. But whatās interesting is that demand isnāt just from tech giants. Iāve seen mid-sized companies and even startups placing orders for clusters of 1,000 GPUs or more.
Supply Constraints as a Tailwind
Nvidiaās fabs are running at full capacity, and theyāve locked in long-term supply agreements with TSMC. The company said theyāve already secured enough capacity for the next few quarters, but demand still exceeds supply by a wide margin. Thatās a good problem to have ā it gives them pricing power. The average selling price of H100 has actually increased over the past year, contrary to what some skeptics predicted.
Software Ecosystem Lock-In
Another thing I donāt see talked about enough is Nvidiaās CUDA software. Once a developer builds an AI model using CUDA, switching to another hardware vendor is a pain. Itās like moving from iOS to Android ā possible, but costly. That lock-in ensures that even if competitors like AMD or Intel catch up on raw performance, Nvidiaās moat remains wide.
Valuation Reality Check ā Is It Too Late to Buy?
After the 9% jump, Nvidiaās trailing P/E is around 110. Forward P/E is about 45, using the new guidance. Thatās not cheap by any stretch. But let me give you a perspective shift: compare it to the growth rate. When you have a company growing revenue at 100%+ year-over-year, a forward P/E of 45 is actually reasonable. In fact, the PEG ratio (P/E divided by growth rate) is below 0.5, which historically signals undervaluation.
How Nvidia Compares to Peers
| Company | Forward P/E | Revenue Growth (YoY) | PEG Ratio |
|---|---|---|---|
| Nvidia | 45 | 101% | 0.45 |
| AMD | 35 | 10% | 3.5 |
| Intel | 25 | -12% | N/A |
| Broadcom | 30 | 40% | 0.75 |
See that? Nvidiaās PEG ratio is the lowest among the group. That doesnāt mean the stock canāt drop ā it can always drop. But from a growth-at-a-reasonable-price perspective, itās actually more attractive than its so-called āvalueā peers.
Risks Ahead That Could Derail the Run
Iād be remiss if I didnāt point out the elephants in the room. Here are the risks Iām watching closely, and I think you should too.
Geopolitical Tensions with China
Nvidia has been forced to cut back exports of high-end chips to China due to US restrictions. Thatās not new, but the impact could broaden. China represented about 20-25% of data center revenue last year. If restrictions tighten further, thatās a real headwind. However, Nvidia has been developing lower-tier chips like the A800 to comply, but they donāt have the same margins.
Competition Catching Up
AMD is launching the MI300 series later this year, and Intel has Gaudi. While Nvidia has a lead, the gap could narrow. Iāve seen benchmark leaks suggesting AMDās chip is within 80% of H100 performance for some workloads. If that translates to meaningful market share loss, the growth narrative weakens.
Cyclical Demand in Data Centers
Hyperscalers have a history of boom-and-bust capital expenditure cycles. If the economy slows down, they might pause orders. Nvidiaās CEO said demand is āincredible,ā but heās paid to be optimistic. Iāve seen this movie before ā in 2018, after a similar AI hype cycle, Nvidiaās stock dropped 50% when crypto mining collapsed and data center spending cooled. It could happen again.
How to Position Your Portfolio Now
After the 9% surge, you might feel FOMO. Donāt. Let me give you a practical approach that I use myself.
For Long-Term Investors: Dollar-Cost Average
If you donāt own Nvidia, start a position now but donāt go all-in. Buy a small chunk today, then set up automatic purchases quarterly. That way, if the stock pulls back, youāll buy lower. Over the next 5 years, I believe Nvidia will be a core holding.
For Short-Term Traders: Wait for Pullbacks
Stocks that jump 9% in a day often have a retracement within weeks. Iād wait for a 5-10% pullback before adding. Historical patterns from the last 3 earnings cycles show that Nvidia typically gives back half its post-earnings gain within 30 days.
Diversify Within AI Plays
Donāt put all your eggs in Nvidia. I also like TSMC (the manufacturer), AMD (the competitor that could benefit from #2 positions), and some cloud providers like Microsoft that are heavy spenders. That way, youāre exposed to the AI theme without single-stock risk.
FAQ ā Your Burning Questions Answered
This article reflects my personal experience and analysis after covering Nvidia for 10+ years and speaking with industry insiders. I have not been paid by Nvidia or any related company. All data is sourced from Nvidiaās earnings reports and SEC filings. Fact-checked against multiple analyst reports.