The Chip Business Is Booming. Why Isn’t Intel?

0:05My colleague Asa Fitch covers a product that's often invisible to the naked eye, but critical to almost everything we do. You should know that some of us on the team kind of refer to you a little bit as Mr. Chips. We know that when there's a chip story, we know who to call. I'll take it. That's a good nickname. I like it. Chips. Those tiny, intricate slices of silicon. Asa calls them the engines of modern life. Just think about, you know, the 80s and the 90s. There were PCs and they had chips. And then there was the smartphone revolution. And then everybody's carrying a smartphone with chips in the smartphone.

0:45And you had cloud computing, you know, these massive data centers, and they had chips. And, of course, now you have AI. And AI sort of runs on chips. So every time that society's advanced, the answer has been, we need some more chips for that. That's been a boon for the companies that make chips. Companies like TSMC, or Taiwan Semiconductor Manufacturing Company, and Samsung. The chipmaker NVIDIA is currently one of the most valuable companies in the world, worth over$3 trillion.

1:22But there's one company that's conspicuously missing from that list of winners. The storied American chipmaker Intel. Intel's stock has fallen over 50 % this year. It's laid off thousands of employees. And last month, it was bumped from the Dow Jones Industrial Average and replaced by NVIDIA. Intel's been in the Dow for 25 years. What a knock for Intel. Why is Intel struggling so much? It's a long story. You could argue it's a tale of hubris, of excessive confidence. It's a tale of rigidity and a lack of flexibility that often large corporations have to contend with.

2:11Welcome to The Journal, our show about money, business, and power. I'm Jessica Mendoza. It's Monday, December 16th.

2:24coming up on the show how intel lost its edge

2:42from the 1980s through the early 2000s intel was everywhere The company became this dominant force with the rise of the personal computer. Everybody needed a computer in their house. Eventually, laptops came along as well. Most of them, almost all of them, had Intel chips inside of them. Intel both designed and manufactured its chips, called CPUs. And it sold them to computer makers like Dell, IBM, and Apple. Pretty much every major computer maker in that era used Intel chips. There was that famous marketing campaign, of course, Intel Inside.

3:26It's like the sticker on the bottom corner of your laptop or your computer or something would say, yeah, Intel Inside in blue. Right, exactly. And, you know, it was a bonanza for a long time.

3:40And then, perhaps because of that success, Asa says Intel started to slip. If you're a dominant player and you're just raking in the money, it's a little bit easier to not be serious about shaking things up. And it goes to the sort of the quote-unquote innovator's dilemma, essentially, where a company becomes a powerful player in one area and then struggles to dominate or make a name for itself in another area. One important area that Intel missed out on was the mobile phone boom in the 2010s. Phones, of course, also need chips But by the time Intel started making chips for phones in a serious way Other companies already dominated And mobile wasn't the only trend that passed Intel by Asus says it also missed out on a pretty big shift in how chips are made Remember, Intel had always designed and manufactured its own chips But increasingly, companies specialized Some companies just designed chips NVIDIA is one of these.

4:52Other companies just manufactured chips, like the Taiwanese company TSMC. So for example, NVIDIA designed chips and then hired TSMC to make them. And TSMC had become extremely successful in this world, was building a growing business around it. Intel, of course, saw that happening. But Intel decided to stick to its guns and not to kind of bifurcate itself in the same way the industry is going. They said essentially, no, we don't want to do that. Instead of specializing, Intel dug in and slipped further behind. You know, the company tries to make the most advanced chips in the world. You know, what that means is like, you know, when you have a chip, it's a piece of silicon.

5:43and upon that silicon, they etch billions upon billions of transistors. And so you might ask, okay, well, what makes one chip better than another? The answer is basically you're able to pack more transistors in that small space. That's very important in the chip world. It's like the most important thing. Yeah. If you can have a lot of transistors, it means you can do a lot of stuff. I mean, you can produce the best AI chips in the world. You can produce the fastest calculating CPUs in the world. Your computer that you're on now can run faster, a lot faster, if it has more transistors. It's like having a bigger brain.

6:24So Intel kind of lost its way in that race. It just started to fall behind these contract chip makers like TSMC in making the most advanced chips. By 2019, things were not going well for the company. That was very clear. There was a sense within the company that, you know, Intel had already sort of lost its glory. Its glory days were over, and the mission was to kind of bring it back. And that's the context into which Pat Gelsinger arrives. Pat Gelsinger, Intel's man with a plan. That's next.

7:29Who is Pat Gelsinger? So Pat Gelsinger is Intel's first chief technology officer, first CTO. So he's a guy who was sort of born and bred Intel. Pat Gelsinger grew up in rural Pennsylvania. He was a farm boy with an associate's degree from a local technical college. But as Gelsinger explained in an oral history, at age 18, his life changed. Intel came recruiting and invited him out to company headquarters for an interview. So, you know, here I am, 18 years old. I've never been on an airplane and I'm giving a free trip to California. According to one Intel interviewer, Gelsinger was, quote, smart, aggressive, arrogant.

8:15He'll fit right in. Gelsinger would spend the next three decades of his career at Intel. You know, when you talk about Intel Inside, that marketing campaign in the 90s, those chips that were going to those computers were, in large part, you know, thanks to Pat Gelsinger. You know, he did a lot in helping design those things. Gelsinger eventually left the company. But in 2021, Intel convinced him to come back as CEO to help turn things around. Gelsinger's plan to do that was bold and risky.

8:52Intel would keep designing and making its chips like it always had. But it would also start manufacturing chips for outside customers. It would directly take on chip manufacturers like TSMC. They were going to go full steam ahead, create this contract chip-making business within Intel, and use that to fill up its factories and do more manufacturing, basically, and make more money. And as part of that, Intel was going to expand its factory footprint, just build more factories. Those factories don't come cheap, by the way. Gelsinger's plan was projected to cost over$100 billion in coming years. Even with outside investment and government money from the CHIPS Act, it would be a hefty bill.

9:41The problem in the chip industry is that you have to make decisions today about what you're going to need several years from now, even five, ten years from now. Gelsinger had to essentially predict many years in advance what the demand might be for these chips and build those factories accordingly. Yeah, you're betting. You're betting. It's fundamentally a bet. So for Intel, given where it was at that time, was it kind of a make or break bet? It was. I mean, it was absolutely. Pat said at the time, you know, it was a bet the company move. And how did that bet go? Intel tried to attract customers for its contract chipmaking business, and it had trouble.

10:29I mean, the big customers for the contract chipmaking business are Apple, NVIDIA, of course, Qualcomm. There are some others. But Intel had this internal goal that they were going to be the second largest contract chipmaker in the world by 2030. And it's almost 2025. And the amount of major customers that Intel was getting in the door was just, you know, it wasn't huge. Chip designers were hesitant to jump manufacturers, especially since Intel didn't have a history as a contract chip maker. And Intel also had a culture issue. Contract chip making is really a customer service business. You know, if your customer wants something, you bend over backwards to make it happen for them.

11:15Right. For Intel, they hadn't really been like that. I mean, their customer had been just Intel. They didn't really have that experience, like dealing with those kinds of customers meant that it was just a little bit more difficult to get that business started. All of this would have been challenging enough, even if the chip industry had stayed relatively stable. But just a year into Gelsinger's massive turnaround effort, one viral product changed everything. Basically, it's ChatGPT. So ChatGPT came along, and that leads to a flurry of interest and investment in generative AI, people developing these big AI models.

11:57And that, in turn, leads to huge investments in chips that are needed to build these kinds of models. And those chips are made almost entirely by NVIDIA. Intel made its name with CPUs, but NVIDIA has always specialized in designing a different kind of chip, called a GPU. It stands for Graphics Processing Unit. GPUs were originally designed to handle the complex graphics in video games. But over the past decade or so, AI researchers discovered another use for them, as the brains of AI. If you want to train or run an AI model, you need a lot of GPUs. And NVIDIA had a huge head start on everyone, including Intel.

12:44After ChatGPT came out, demand for NVIDIA's chips exploded, and it's kept up ever since.

12:52So that was a huge, huge challenge for Intel. When the AI boom really took hold, everything was going to AI chips, and significantly less amount of money was going to Intel's chips. So there was this dramatic shift that really hurt Intel financially at a time when the company really just needed money to build these huge factories that cost billions upon billions of dollars and execute this turnaround plan that Pat had concocted. So it really could not have come at a worse time for Intel. Intel was burning through cash to get its chip making business off the ground. And with the AI boom cutting into its sales, it was cash the company increasingly didn't have.

13:38So Gelsinger and his team economized. Intel laid off thousands of staff and suspended its dividend. It cut billions of dollars in costs. But the fundamental picture didn't change. Since Gelsinger took the helm, Intel's stock has lost 60 % of its value. Its sales have plummeted. And in October, the company posted the biggest quarterly loss in its history. Then, last week... Intel just moments ago announcing that CEO Pat Gelsinger retired effective yesterday. Pat Gelsinger is retiring from the Santa Clara company. He stepped down from the board of directors effective immediately. According to someone familiar with what happened...

14:19It was retirement, but the circumstances were that the board gave him the option to retire or be removed, and he chose to retire. And what was the reaction in your world to that news? You know, many people were surprised by Gelsinger being ousted because, you know, this is a very ambitious turnaround plan. The company had already invested a lot of resources in making the turnaround work. Intel has been on successive rounds of cost-cutting for the past couple of years at this point. They've done layoffs. They've cut costs wherever they can. and all of that has been in service of making this plan work.

15:00And so there's a lot of surprise, actually, that at this juncture, Gelsinger would be ousted. But, you know, I think the board just sort of lost confidence in Pat and the plan and the speed at which it was going and the results from it. In a statement, Gelsinger said that leading Intel has been, quote, the honor of my lifetime. He called his departure bittersweet. An Intel spokesman said the changes the company had made over the past few years had revitalized its chip technology and laid the groundwork for the company's future. Asa says that Intel may ultimately stick with Gelsinger's turnaround strategy, just under a new chief executive.

15:41Last week, the company's interim CEOs said Intel is still highly invested in building out a contract chip-making business.

15:53looking at sort of intel's journey here when we think about other sort of storied american companies that might be watching what would you say is the cautionary tale here it's hard to know how this period of the company's history will be going to be seen but it may be that it's seen as you know no ceo could have turned around intel But if somebody could have, it would have been Pat Gelsinger, and this would have been the plan. It just didn't work out. And maybe the history shows that in the final analysis that if Intel were to be set or right again, it would have had to have happened a long time before Pat Gelsinger took over as CEO.

16:36When you get that big, it's harder to, you know, as Lou Gerster wrote in his book about IBM, make elephants dance, if you will. You know, make large companies do agile things is hard. And what Intel has needed in the past years has been agility.

17:07That's all for today, Monday, December 16th. The Journal is a co-production of Spotify and The Wall Street Journal. Additional reporting in this episode by Ben Cohen.

17:22Thanks for listening. See you tomorrow.