Qualcomm Sold Server Chips Once Before and Quit Within a Year. History Says What Has to Be Different This Time.

Qualcomm (NASDAQ:QCOM) said on Tuesday that Amazon (NASDAQ:AMZN) could buy up to $60 billion of its server chips and related technology over the next decade. The two companies plan to co-develop multiple generations of customized silicon for Amazon Web Services' data centers, starting with chips built for AI (artificial intelligence) inference (the day-to-day work of running already-trained AI models). Shares of the chipmaker rose 3% on the announcement.

But Qualcomm has been here before. In November 2017, the company announced commercial shipment of its Centriq 2400, billed as the world's first 10-nanometer server processor.

Within six months, Bloomberg reported that the company was preparing to give up on the business. By the end of 2018, Qualcomm was cutting the data center group from a peak of more than 1,000 people to about 50.

What has to be different this time?

Image source: Getty Images.

Centriq never found a committed buyer

The 2017 failure wasn't a story of weak engineering or half-hearted effort. Qualcomm's data center chief at the time, Anand Chandrasekher, called the launch "the culmination of more than four years of intense design, development and ecosystem enablement effort."

The 48-core version of the chip listed at $1,995, and Qualcomm claimed more than 4 times the performance per dollar of Intel's highest-end server processor. Microsoft and other cloud providers demonstrated workloads on it at the launch event.

What no one did was commit to buying it in volume. Demonstrations never became the kind of orders a company can build a business on.

By May 2018, Bloomberg was reporting that Qualcomm was exploring a shutdown or a sale of the unit, part of a push to cut $1 billion in annual costs.

And Amazon, notably, chose to build rather than buy. It introduced its own Arm-based Graviton server processors in November 2018, just weeks before Qualcomm's retreat became official.

Amazon's commitment is in writing

The new agreement addresses that failure directly. As part of the deal, Qualcomm issued Amazon warrants to buy 25 million of its shares at $161.26 apiece. According to the company's filing, those shares vest in tranches tied to the execution of commercial agreements, the placement of binding purchase orders, and actual purchases of Qualcomm's server chips, systems, and services. The payments that count toward vesting are capped at $60 billion through 2036.

In other words, Amazon gets the full value of its warrants only if it buys. That structure, I'd argue, is the single biggest difference from 2017.

And some buying is already locked in. Of the 25 million warrant shares, 3.75 million vested immediately, based on what the filing calls initial purchase commitments.

To be fair, the $60 billion is a ceiling on what counts toward vesting, not a promised spend. But Centriq died without a committed dollar. This attempt starts with purchase commitments already made -- and a partner that profits by spending more.

This time, the roadmap is the strategy

The other thing Centriq lacked was a parent willing to fund it through the slow years. The 2018 exit was a cost cut as much as a verdict on the product. Today, the data center business sits at the center of Qualcomm's plans.

Meta Platforms agreed in June to use the company's data center processors across multiple generations, starting with the Dragonfly C1000, expected in 2028. Management targets more than $15 billion in data center revenue by fiscal 2029, and CEO Cristiano Amon expects non-handset revenue growth to accelerate from 24% in fiscal 2026 to more than 60% in fiscal 2027 -- a sharp acceleration, if it lands.

Qualcomm likely needs it to land. In the fiscal third quarter of 2026 (the period ended June 28, 2026), handset chip revenue fell 20% year over year to $5.1 billion, and total revenue slipped 4% to $9.9 billion. Automotive revenue rose 61% year over year, and internet of things revenue grew 9%. But the two combined remain smaller than the handset business they are meant to offset.

Ultimately, this attempt has the committed anchor customer the first one never found, and a roadmap the company is funding as its main growth bet. What the terms can't settle is execution. The $60 billion is a cap, not a guarantee, and the revenue ramp isn't expected before fiscal 2027.

Shares trade around $176 as of this writing, or about 17 times the earnings analysts expect for next fiscal year -- hardly a price with data center success already baked in. Still, I'd stay on the sidelines for now. I want to see those purchase orders show up as revenue first.

Daniel Sparks and his clients do not have positions in any of the stocks mentioned. The Motley Fool has positions in and recommends Amazon, Intel, Meta Platforms, Microsoft, and Qualcomm. The Motley Fool has a disclosure policy.