Is this 1999 all over again??

Semiconductor stocks are going absolutely parabolic right now.

In just the last month:

• Intel (INTC) is up 105%.

• Qualcomm (QCOM) is up 88%.

• Micron (MU) is up 83%.

• SanDisk (SNDK) is up 79%.

These are absurd moves in just four weeks.

Now naturally, investors are asking the obvious question:

Is this another bubble?

The answer is… sort of.

But not in the way most people think.

Here’s Why Chip Stocks Are Exploding Higher

This rally isn’t happening because investors randomly decided to gamble on semiconductors.

There’s a real reason behind it.

Artificial intelligence has created an enormous demand boom for chips.

Every AI chatbot, AI image generator, self-driving system, and data center requires massive amounts of computing power.

And chips are the foundation that makes all of that possible.

That means companies like Microsoft, Amazon, Meta, Google, and OpenAI are now spending hundreds of billions of dollars building AI infrastructure.

And nearly all of that money eventually flows downstream to semiconductor companies.

That’s why these stocks keep going higher.

The businesses themselves are suddenly making far more money than Wall Street expected a year ago.

Revenue is accelerating.

Profit margins are expanding.

And earnings estimates keep getting revised higher.

That combination is rocket fuel for stocks.

Should You Be Blindly Buying Chip Stocks?

This is where investors need to be careful.

Because not all chip stocks are the same.

In other words, saying “I’m bullish on semiconductors” is kind of like saying “I’m bullish on technology.”

Okay… but which part?

Let’s quickly break down the major categories.

1. GPUs & AI Accelerators – AI “Training” Chips

Before AI can answer a question, it first has to be trained.

That means feeding models massive amounts of data so they can learn patterns and relationships.

This process requires enormous computing power, which is why GPUs have become so valuable.

Unlike traditional chips that handle tasks one at a time, GPUs can process millions of calculations simultaneously.

Nvidia (NVDA): Nvidia remains the undisputed king of AI training chips with roughly 90% market share. Its CUDA software ecosystem has become deeply embedded across the AI industry, making it extremely difficult for competitors to catch up despite the stock’s massive run. Current Forward P/E: 19x.

AMD (AMD): AMD has emerged as Nvidia’s primary challenger by offering powerful AI GPUs with large memory capacity designed for massive AI models. The company continues gaining traction with hyperscalers looking for cheaper alternatives to Nvidia’s increasingly expensive hardware. Current Forward P/E: 36x.

Google TPU & Amazon Trainium: Unlike Nvidia and AMD, these are custom in-house AI chips designed specifically for internal cloud infrastructure. The biggest advantage is cost. Both Google and Amazon are trying to lower AI computing expenses by reducing reliance on Nvidia’s premium-priced GPUs.

2. CPUs – The “Everyday AI” Chips

Training an AI model is only the first step.

After it’s trained, the AI then has to perform real-world tasks billions of times per day. That process is called “inference.”

This includes everything from ChatGPT responses and AI assistants to recommendation algorithms, cybersecurity systems, factory automation, and self-driving technology.

This is where CPUs become important.

AMD (AMD): AMD is the performance leader in CPUs, continuing to take share in data centers. This makes the company one of the largest beneficiaries of AI models moving from the training phase to inference. Current Forward P/E: 36x.

Intel (INTC): Intel is more of a turnaround story at this point. Investors are betting the company can rebuild its manufacturing leadership and gain some sort of relevance in the AI buildout.  The stock has rallied hard, but valuation has also become stretched relative to the company’s current growth profile. Current Forward P/E: 85x.

Qualcomm (QCOM): Qualcomm is becoming one of the biggest early winners in “AI on-device” computing, with chips designed to run AI directly on laptops and smartphones without relying entirely on cloud servers. Current Forward P/E: 22x.

3. Memory & Storage – The “AI Fuel”

An AI system is useless if it can’t quickly access or store data.

That’s where memory and storage chips come in.

As AI models become larger and more complex, demand for these chips is exploding.

That said, investors should remember that memory has historically been one of the most cyclical areas of semiconductors.

Periods of tight supply and soaring prices often attract massive new production capacity, which leads to oversupply and plummeting prices later on.

• Micron (MU): Micron has become one of the biggest AI winners because it produces HBM (High Bandwidth Memory), which helps AI chips access data incredibly quickly. Without it, even the world’s fastest GPUs would slow down waiting for information. Current Forward P/E: 8x.

• SanDisk (SNDK): AI systems also create enormous amounts of data that need to be stored somewhere. That’s driving strong demand for high-speed storage products like SSDs and enterprise data center drives. Current Forward P/E: 9x.

So… Is This 1999 All Over Again?

Parts of the semiconductor market are clearly overheated.

Stocks moving 80%-100% in a month is not normal.

At some point, many of these names will have sharp pullbacks. That’s just the reality of moves like this.

But there is one major difference between today and the dot-com bubble:

Back then, investors were mostly betting on ideas.

Today, these companies are generating enormous revenue, massive profits, and real cash flow from a very real AI infrastructure boom.

That said, investors still need to be selective.

Some parts of the semiconductor industry are much better businesses than others.

Memory, for example, has historically been one of the most volatile and cyclical areas of the market.

These stocks can soar during shortages and collapse once supply catches up.

And valuations still matter too.

Intel trading around 85x forward earnings makes for a very different risk/reward setup than a company growing rapidly at a much cheaper valuation.

Personally, Nvidia remains my favorite way to play the AI buildout long term.

Despite the massive rally, the stock still trades around 19x forward earnings because earnings growth has exploded so quickly.

So to answer the question more directly – no, I don’t think this is 1999 all over again.

AI is driving one of the largest infrastructure buildouts we’ve seen in decades, and semiconductors are sitting directly at the center of it.

The key now is separating the companies building durable long-term businesses from the ones simply riding the excitement.