Intel’s deal with Amazon might be just the Hail Mary it needs

Intel has not been having a great year in 2024, with its stock price down 54% year-to-date compared to a 19% gain by the S&P 500, and its last earnings report provided a perfect summary of the company’s recent struggles. To quote myself from August:

Intel reported lackluster earnings last week, with a 1% decline in year-over-year revenue and a $1.61 billion operating loss, including a $2.8 billion loss stemming from its Foundry unit that generated $4.3 billion in revenue (4% year over year growth). Even worse, the company stated that it was slashing 17,500 jobs and suspending its dividend just five months after announcing that the CHIPS Act funding would create almost 30,000 jobs.

One of Intel’s problems is that it has repeatedly missed deadlines for releasing more powerful chips, causing it to fall behind competitors like Nvidia and AMD in the AI arms race. Another issue has been the company’s struggle to build a large customer base for its foundry business. 

Foundries manufacture chips that were designed by other companies, and TSMC dominates the foundry market, with data from Statista showing that it has ~62% of the foundry market share, with its largest competitor, Samsung, only holding an 11% market share.

One reason for TSMC’s success is that its 3nm chips are the most advanced technology on the market. Another TSMC advantage, however, is its lack of conflicts of interest. In 1987, TSMC was founded as the world’s first dedicated semiconductor foundry company, and it doesn’t design its own chips. Companies simply send TSMC their designs and pay them to produce chips.

While Intel has grand foundry ambitions, it also designs and sells its own chips, which created an inherent conflict of interest. Investor and technology analyst Kevin Xu explains it well here:

As a customer, how can you be certain that Intel will prioritize manufacturing your chips over its own? To address these concerns, Intel announced in October 2022 that it would “create greater decision-making separation between its chip designers and chip-making factories as part of Chief Executive Pat Gelsinger’s bid to revamp the company and boost returns.” For the last two years, we have waited to see if this move would attract big-name customers, and on Monday, we got our answer:

In the same 24-hour period, Intel announced that it was turning its foundry business into a “wholly owned subsidiary,” making it totally operationally independent from the rest of the company, and it signed a “multibillion-dollar agreement for Amazon.com’s cloud-computing arm to manufacture chips at Intel factories using an advanced chip-making technology expected to go into production next year.”

There is a common phenomenon in the venture market where investors might hesitate to invest in a startup until a big-name fund like a16z or Sequoia writes a check, then everyone wants to participate in the next funding round. Right or wrong (as we saw with FTX), a well-known fund investing in a startup is a positive signal to the market, giving other investors more trust in the company.

Amazon may be Intel’s Sequoia: if the $2 trillion tech giant is willing to invest in Intel, other companies might do the same. To be clear, Intel is still in a hole: Intel Foundry lost $2.8 billion last quarter, and management noted that foundry investments would continue to weigh on its operating profits through the end of the year. However, Amazon has provided some much-needed positive momentum for the ailing chipmaker.

Also, if you happen to believe that the The Economist cover is really a contrarian indicator, things are looking good for Intel now: