For decades, Intel was the world’s most valuable chipmaker.

Now it is one-thirtieth the size of Nvidia, one-eighth the size of Broadcom, and roughly half the size of AMD, Qualcomm, and Texas Instruments.

In the last few days, the company’s fall from grace has prompted Qualcomm to float merger talks and private equity firm Apollo to reportedly offer $5 billion to shore up the balance sheet.

So how did the storied 56-year-old company get to this point?

Intel’s Spiral Is A Decade In The Making

To understand Intel’s downward spiral, it’s important to understand that Intel primarily produces Central Processing Units (CPUs).

Think of a CPU as the head chef in a kitchen.

The chef is really good at handling complex tasks one by one, like carefully crafting a dish.

The chef focuses on making sure everything is done perfectly, but can only do so much at once.

These CPUs are used to power PCs and laptops. Previously, Intel dominated this market.

That was until Intel’s 7-nanometer and 10-nanometer chips were significantly delayed in the late 2010s and early 2020s.

This opened the door for AMD to steal significant market share.

Whereas AMD barely had market share in server CPUs a decade ago, Mercury Research estimates AMD had about 25% of the market as of Q1 2024 – a major blow to Intel’s core business.

The “New” Intel

In the chip world, most companies don’t actually manufacture their own chips.

Nvidia, AMD, Qualcomm, Broadcom and many more use what’s known as the “fabless” model, where they outsource manufacturing to companies like TSMC and Samsung.

Intel is unique where they both design and manufacture their own chips.

When long-time Intel employee Pat Gelsinger became CEO in early 2021, he decided to double-down on manufacturing.

It was an ambitious bet, but the stage seemed to be set in their favor. Rising manufacturing demand has significantly increased wait times at existing manufacturers TSMC and Samsung. Plus, this strategy aligns with the onshoring trend we’ve seen in the aftermath of Covid-19, removing reliance on countries like China and Taiwan (where TSMC is located).

Gelsinger initially planned to acquire companies to build out Intel’s manufacturing capabilities.

Unfortunately, acquisitions of GlobalFoundries and Tower Semiconductor failed which forced Intel to build out its manufacturing capabilities the old-fashioned way – by spending billions of dollars over many years constructing these highly technical facilities.

You can see the double-edged sword here: at the exact moment Intel was bleeding market share in its core CPU division, it began racking up billions of dollars in costs to expand its manufacturing business.

The company’s financials certainly took a hit. Both revenue and earnings have declined substantially since the plan was enacted.

Intel Missed The AI Revolution

Occurring in tandem with Intel’s core business losing market share and the very expensive manufacturing buildout was the fact the company completely underinvested for the AI revolution.

While CPUs are a head chef in the kitchen, Graphics Processing Units (GPUs) are like a team of line cooks, and they’re what is powering AI.

Instead of handling one task at a time, the line cooks can handle many smaller tasks all at once – like chopping vegetables or grilling burgers.

They're fast and efficient at doing repetitive tasks simultaneously, which is why GPUs are great for things like graphics and AI.

Nvidia dominates this red-hot market, which is why that stock is up 135% this year.

Intel is down 55% this year and currently trades around $23/share.

According to Gelsinger, however, this year is supposed to be the “trough” when all of these negative factors converge and the worst gets put in the rearview.

As an investor, I support the manufacturing strategy.

It’s a “pickaxe and shovel” type playbook during the early stages of AI that also aligns with the onshoring trend we’ve seen in the aftermath of Covid-19.

However, it will be years before Intel can compete with TSMC and Samsung.

Until the company can show returns on the billions of dollars they’re currently investing and stop the bleeding in its core CPU business, expect the stock to continue trading wildly.

Before that can happen, Gelsinger first needs to weigh his options regarding the recent interest from Qualcomm and Apollo.