Moore’s Flinch

First, Apple gave up on Intel. Now, Intel might be giving up on Intel.

In one of the wildest Friday-night news dumps in a while (in a week full of absolutely insane news stories), word out of the Wall Street Journal is that Qualcomm, a semiconductor giant whose prominence has grown on the back of the smartphone, was interested in buying Intel outright.

Even a year ago, a story like this would have seemed surprising or unlikely. But Intel has been in a major state of transition and has run into serious financial struggles. The problem is essentially the result of a significant business reshuffling that is leading to major losses, even as many of its individual businesses are profitable.

There has been a lot of disruption in the Intel camp of late, in part because it missed some important trends and allowed its competitors to gain serious ground. To give you an idea, I’ll turn it over to reporters Lauren Thomas, Laura Cooper, and Asa Fitch:

For years, Intel was the biggest semiconductor company in the world by market value, but it now lags behind rivals including Qualcomm, Broadcom, Texas Instruments and AMD.

In August, following a dismal quarterly report, Intel said it planned to lay off thousands of employees and pause dividend payments as part of a broad cost-saving drive.

Gelsinger last month laid out a roadmap to slash costs by more than $10 billion in 2025, as the company reported a loss of $1.6 billion for the second quarter, compared with a $1.5 billion profit a year earlier.

Then, director Lip-Bu Tan abruptly stepped down from Intel’s board in late August. His departure was a shock to many who saw him as an eventual leader of part of Intel’s business, should it be split up.

Intel is sort of in this weird between-rock-and-hard-place transition point, where it sees the future of its business as a chip-fabricator-for-hire, along the lines of TSMC (Taiwan Semiconductor Manufacturing Company Limited). (It just scored a big win by signing on Amazon.)

The latest edition of the Microsoft Surface Laptop, one of the first mainstream laptops with a Qualcomm Snapdragon processor. (via Microsoft)

TSMC’s model has allowed companies like Apple and AMD to leapfrog Intel from an innovation standpoint in recent years, in part because of its work in winnowing down process nodes. Intel, which usually fabricates its own chips, has had to use TSMC in recent years, which seems like it’d be quite humbling.

Like Microsoft, Intel basically missed the boat on the mobile revolution, and despite aggressive branding of its recent chips to play up AI, they’re not anyone’s first choice for a GPU or NPU. Owning a fab-for-hire could put it in a position where other massive companies are reliant on it, just as they are TSMC. But it hasn’t worked out that way thus far, because what they’re trying to do is comically expensive.

The press picture around Intel has been messy throughout 2024—notably, their most powerful consumer CPUs have had a show-stopping bug that the company has struggled to squash. But ultimately, the real problem is that the company is trying to build for the next stage of its business under CEO Pat Gelsinger, but that’s proving really complicated. Rumors are flying, and the Qualcomm rumors are just the latest in a long line.

The fact that Intel was reportedly considering splitting off its chip assets in favor of the fabrication arm reminds me of a similar breakup of a giant technology company, Hewlett-Packard. About a decade ago, after a failed tablet and a failed acquisition, the company decided to split off its consumer business, the one that makes printers and laptops, as HP, while putting its enterprise assets—the remnants of Itanium, the server gear, etc.—into a new company called HPE. The laptop-maker is the bigger of the two, but the enterprise arm is no slouch. One could see Intel making a similar split.

The reason for Qualcomm wanting to buy Intel is likely complex, but it still feels like the simple explanation has value: If Qualcomm buys Intel, the transition to ARM chips on the Windows side, which likely would take a decade to fully take effect and would have two large competitors pushing back at every turn, gets cut by half a decade. Sure, AMD would still be there, but one gets the feeling that AMD would love to make ARM chips itself someday. And since AMD is a fabless chip designer just like Apple and Nvidia, meaning they outsource the chip manufacturing to TSMC and similar firms, they likely would have an easier transition than Intel might.

And Qualcomm has a big enough imprint right now, thanks to their scale in mobile, that they might just benefit from having Intel’s fabrication resources in-house.

There are a lot of questions to be answered—would antitrust regulators let this happen, even given a wounded Intel? If Qualcomm buys, does Intel exist as a standalone arm of Qualcomm? And what about the 45 years of existing PCs dependent on x86 technology?

If it happens, it would be a sad loss. The PC industry as we know it wouldn’t exist without Intel. But maybe nothing is permanent. Not even the company that made Moore’s Law the best-known theorem in computing.