You've seen the charts. They look vertical. It's the kind of growth that makes seasoned Wall Street vets sweat and retail investors check their brokerage apps every fifteen minutes. People keep asking the same question: will nvidia keep going up or are we just watching the world’s most expensive firework display?
Honestly, there isn't a simple yes or no. If anyone tells you they know for sure, they’re probably trying to sell you a newsletter. But we can look at the plumbing of the global economy. Right now, Nvidia isn't just a chip company; it's the toll booth for the entire artificial intelligence revolution.
The GPU Shortage That Isn't Actually Over
Back in 2023, you couldn't get an H100 chip if you were royalty. Now, supply has loosened up a bit, but the demand from hyperscalers—think Microsoft, Google, and Meta—is still borderline voracious. Zuckerberg alone is dropping billions to stack hundreds of thousands of these units.
Why? Because training a Large Language Model (LLM) isn't a "one and done" situation.
💡 You might also like: How to Actually Use a Z Table Standard Normal Distribution Without Getting a Headache
Every time OpenAI or Anthropic wants to leap from one version to the next, the compute requirements don't just grow. They explode. We’re talking exponential scaling. If you want a smarter chatbot, you need more silicon. If you want a robot that can fold laundry, you need even more. Nvidia’s Blackwell architecture is the next logical step here, promising massive jumps in performance that make the previous generation look like a pocket calculator.
Is the AI Bubble About to Pop?
Critics love the word "bubble." They point to the dot-com crash of 2000 and say Nvidia is the new Cisco. It’s a fair comparison on the surface. Cisco built the routers that built the internet, then the stock crashed when the build-out finished.
But there’s a nuance here most people miss.
Cisco’s hardware was somewhat commoditized. Nvidia has CUDA. That’s their secret sauce. It’s a software layer that millions of developers have spent a decade learning. Switching from Nvidia to a competitor like AMD or an in-house chip from Amazon isn’t just about swapping a card; it’s about rewriting millions of lines of code. That’s a "moat" made of steel and concrete.
Still, the risk is real. If companies like Meta and Salesforce don't start seeing a massive Return on Investment (ROI) from their AI spending, they’re going to stop buying chips. Eventually, shareholders want to see profits, not just "cool" demos of AI agents that can write poems about brunch.
📖 Related: Why The Night Only Lives Today is Redefining Digital Ephemerality
The Geopolitical Elephant in the Room
We have to talk about Taiwan. It’s the single biggest point of failure for the "will nvidia keep going up" thesis.
TSMC manufactures Nvidia's designs. If anything happens to the peace in the Taiwan Strait, the global tech economy doesn't just dip—it stops. Nvidia is trying to diversify, looking at Intel’s foundries or US-based plants, but that’s years away.
Then there are the export bans. The US government is constantly tightening the screws on what Nvidia can sell to China. China used to be a huge chunk of their revenue. Jensen Huang is clever; he designs "nerfed" versions of chips to stay under the regulatory limit, but it’s a constant game of cat and mouse with the Department of Commerce.
Why Blackwell Changes the Math
If you're looking for a reason the stock might keep climbing, look at the energy problem. AI eats electricity like nothing we've ever seen. Data centers are straining power grids from Virginia to Ireland.
✨ Don't miss: Cost of iPhone SE: What Most People Get Wrong About Apple's Budget King
Nvidia’s new Blackwell chips aren't just faster; they’re more energy-efficient per flop. In a world where power is the bottleneck, the company that provides the most "intelligence per watt" wins. This isn't just about graphics anymore. It's about global energy infrastructure.
The Competitors Are Catching Up (Sorta)
AMD’s MI300 series is actually quite good. Intel is desperate to get back in the game with Gaudi. Even Groq is making noise with their LPU (Language Processing Unit) architecture.
But Nvidia has a head start that feels almost unfair.
They have the best engineers, the most cash, and a CEO who sleeps in a leather jacket and thinks in decades. Most importantly, they have the ecosystem. It’s like trying to start a new social media platform when everyone you know is already on Instagram. You might have better features, but if the people aren't there, it doesn't matter.
The Valuation Headache
Let’s be real: Nvidia is expensive.
By any traditional metric—Price-to-Earnings (P/E), Price-to-Sales—it looks like a nosebleed. But "expensive" is relative. If they continue to grow earnings by 100% or 200% year-over-year, that P/E ratio "compresses" very quickly. You’re betting on the future, not the present.
If the AI narrative shifts from "training" models to "inference" (running the models), the market might change. Inference doesn't always require the most expensive, high-end H100s. It can be done on cheaper chips. If the world moves toward smaller, local AI models on our phones and laptops, Nvidia might lose some of its iron grip on the data center.
What to Watch Next
Keep an eye on the "Big Tech" earnings calls. Don't look at their profits; look at their Capital Expenditure (CapEx). If Microsoft says they are increasing their spending on AI infrastructure, that’s basically a direct deposit into Nvidia’s bank account.
Also, watch the software side. We need a "killer app" for AI that isn't just a search engine or an image generator. We need something that drives massive corporate productivity. If that happens, the sky is the limit.
Actionable Steps for Investors
- Don't FOMO in at the all-time high. Even the best stocks have 20% pullbacks. If you believe in the long-term story, consider dollar-cost averaging rather than dumping everything in at once.
- Watch the margins. If Nvidia has to start cutting prices to compete with AMD, the "infinite growth" story ends.
- Diversify into the "pick and shovel" plays. If you think Nvidia is too risky, look at the companies providing the cooling systems for data centers or the electrical equipment for the power grid.
- Read the 10-K filings. Seriously. Look at the "Risk Factors" section. It’s the only place where the company is legally required to tell you why they might fail.
The question of whether will nvidia keep going up ultimately depends on your time horizon. In the next three months? It could go anywhere. In the next ten years? You're betting on whether AI will be as big as the internet. If you think it is, Nvidia is the backbone of that reality. Just keep your seatbelt fastened, because the volatility isn't going away anytime soon.
Pay attention to the quarterly "beat and raise" cycles. Nvidia has a habit of setting the bar high and then clearing it by a mile. The moment they only "meet" expectations instead of crushing them, the market will likely overreact. That's usually where the best opportunities—or the biggest warnings—hide.