Pat Gelsinger is gone. His successor, Lip-Bu Tan, has barely settled into the corner office. And already Intel is making moves that would have seemed improbable even a year ago — forging partnerships with SpaceX, Tesla, and xAI that place the struggling chipmaker at the center of Elon Musk’s sprawling industrial ambitions.
The announcement, which surfaced in early April 2025 and was first reported by The Motley Fool, confirmed that Intel Foundry Services — the company’s contract chip-manufacturing division — has secured agreements to produce semiconductors for SpaceX, Tesla, and Musk’s artificial intelligence venture xAI. The deals represent a significant vote of confidence in Intel’s foundry ambitions at a moment when confidence is exactly what the company needs most.
But here’s the question Wall Street is really asking: Is this enough?
Intel’s stock has been battered. The company lost more than 60% of its market value over the past two years. Revenue has contracted. Its once-dominant position in PC and server processors has eroded under sustained pressure from AMD, Nvidia, and an increasingly capable ARM architecture pushed by Apple, Qualcomm, and others. The foundry business — which Intel has positioned as its path to relevance in the next decade — has been hemorrhaging cash, posting operating losses exceeding $7 billion in 2024 alone.
Against that backdrop, landing Musk’s companies as customers isn’t just a commercial win. It’s a narrative shift.
Why Musk’s Companies Need Intel — and Vice Versa
The logic of the partnership runs deeper than headlines suggest. SpaceX, Tesla, and xAI each have enormous and growing semiconductor needs, and each has reasons to diversify away from exclusive reliance on TSMC, the Taiwanese giant that fabricates chips for nearly every major technology company on earth.
SpaceX launches more rockets than any entity in history. Its Starlink satellite constellation — now numbering more than 6,000 satellites in orbit — requires custom chips designed for the harsh environment of space: radiation-hardened processors, specialized communications silicon, and power-efficient computing modules. SpaceX has historically designed much of this silicon in-house and relied on a mix of fabrication partners. Adding Intel to that roster gives SpaceX supply chain redundancy at a time when geopolitical tensions around Taiwan make single-source dependencies increasingly uncomfortable for defense-adjacent companies.
Tesla’s chip appetite is similarly voracious. The company designs its own Full Self-Driving (FSD) inference chips, currently manufactured by Samsung. But Tesla’s next-generation hardware — including the AI training chips being developed for its Dojo supercomputer program — could benefit from Intel’s advanced packaging technologies and its 18A process node, which Intel claims will be competitive with TSMC’s most advanced offerings by late 2025 or early 2026.
Then there’s xAI. Musk’s AI company, which built one of the largest GPU clusters in the world at its Memphis data center, is racing to train models that compete with OpenAI’s GPT series and Google’s Gemini. Custom silicon — application-specific integrated circuits, or ASICs — is the logical next step for any AI company looking to reduce its dependence on Nvidia’s expensive and supply-constrained GPUs. Intel Foundry could manufacture those custom AI accelerators.
For Intel, the calculus is straightforward. The foundry business needs anchor customers. Big ones. Loudly announced ones. The kind that make other potential customers — companies that have been skeptical about Intel’s manufacturing capabilities — sit up and reconsider.
Microsoft was the first major external foundry customer Intel announced, back in 2024. Adding three Musk-affiliated companies in one stroke dramatically expands the customer list and, more importantly, signals that serious engineering organizations with demanding specifications trust Intel’s manufacturing roadmap.
That trust has been hard-won. Intel’s 18A process node has been in development for years, and the company has repeatedly pushed back timelines. Industry analysts have questioned whether Intel can actually deliver competitive yields — the percentage of functional chips produced from each silicon wafer — on its most advanced nodes. Poor yields mean higher costs and lower margins, a death spiral for a foundry business trying to compete with TSMC’s manufacturing discipline.
But recent reports suggest Intel has made meaningful progress. The company has been sharing test chip data with potential customers, and the Musk deals imply that at least some of those results were convincing enough to secure commitments.
The CHIPS Act Lifeline and the Geopolitical Angle
None of this happens in a vacuum. The U.S. government has committed approximately $8.5 billion in direct subsidies to Intel under the CHIPS and Science Act, plus billions more in loans and tax incentives. The explicit goal: rebuild domestic semiconductor manufacturing capacity so the United States isn’t dependent on Taiwan — a 110-mile-wide island across the strait from an increasingly assertive China — for the chips that power everything from smartphones to missile guidance systems.
Intel is the single largest beneficiary of the CHIPS Act. And the government’s investment thesis depends entirely on Intel Foundry becoming a viable, world-class contract manufacturer. Every new customer Intel signs reinforces the argument that taxpayer money is being well spent. SpaceX, with its deep ties to the Department of Defense and NASA, is a particularly compelling name on that customer list.
The geopolitical dimension extends further. The Trump administration has escalated tariff actions against Chinese technology imports and tightened export controls on advanced semiconductor equipment. This environment creates both risk and opportunity for Intel. Risk, because tariffs increase costs across supply chains. Opportunity, because the political pressure to onshore chip manufacturing has never been greater — and Intel is the only American company with the infrastructure and ambition to operate leading-edge fabs on U.S. soil.
TSMC is building factories in Arizona, yes. Samsung is constructing a fab in Texas. But both are foreign-headquartered companies, and political winds could shift in ways that favor a domestic champion. Intel is positioning itself as that champion.
So the Musk partnerships serve multiple masters: commercial revenue, geopolitical narrative, and government relations. A tidy convergence of interests.
Still, skeptics abound. And they have data on their side.
Intel Foundry’s $7 billion operating loss in 2024 was staggering. The division has been described internally as a startup bolted onto a legacy company — one that must simultaneously serve Intel’s own product divisions (which have historically received priority) and attract external customers who are, by definition, Intel’s competitors in some markets. That tension hasn’t been fully resolved.
Lip-Bu Tan, who took over as CEO after Gelsinger’s departure in late 2024, has signaled a willingness to make hard choices. He’s reportedly considering structural separation of the foundry business — potentially spinning it into a distinct subsidiary with its own governance, or even bringing in outside investors to provide additional capital and credibility. The Musk deals could accelerate that process by demonstrating that external customers see Intel Foundry as a legitimate standalone entity, not just a captive manufacturing arm.
There’s also the question of volume. Announcements are not revenue. The financial impact of these partnerships will depend on the specific chips being manufactured, the process nodes involved, and the production volumes ramping over time. Early-stage foundry relationships often begin with test chips and small production runs before scaling. It could be 2027 or later before these deals contribute meaningfully to Intel’s top line.
What This Means for the Competitive Map
The broader semiconductor industry is watching closely. TSMC remains the undisputed leader in contract chip manufacturing, with roughly 60% global market share in advanced nodes. Samsung Foundry is a distant second. Intel Foundry is third — or arguably still in the process of earning that ranking.
But the competitive dynamics are shifting. TSMC’s dominance, while formidable, creates concentration risk that large customers increasingly want to mitigate. Apple, Nvidia, AMD, and Qualcomm all depend on TSMC. If a natural disaster, military conflict, or even a sustained power crisis hit Taiwan, the consequences for the global technology industry would be catastrophic. Diversification isn’t just prudent. It’s becoming a boardroom imperative.
Intel doesn’t need to match TSMC’s scale or yield perfection to capture meaningful foundry share. It needs to be good enough — at competitive enough pricing — to serve as a credible second source for customers who can’t afford single points of failure. The Musk deals suggest Intel is crossing that threshold for at least some use cases.
And Musk himself brings an intangible asset: attention. When Elon Musk is involved, the world notices. Every engineer at every chip design company will register the fact that SpaceX, Tesla, and xAI chose Intel Foundry. That awareness alone has commercial value, even before a single wafer ships.
For Intel’s stock, the partnerships provide a floor of optimism beneath what has been a relentlessly negative narrative. The shares remain deeply discounted relative to their 2021 highs. Analysts are split: some see a generational buying opportunity in a company with irreplaceable assets (fabs, process technology, x86 architecture, government backing), while others see a value trap — a company whose best days are behind it, burning cash on a foundry strategy that may never achieve profitability.
The truth, as usual, probably lies somewhere in between. Intel has real technology, real factories, and now real customers for its foundry business. But execution risk remains enormous. The 18A node must deliver on its promises. Yields must improve. The foundry must eventually turn profitable. And Intel’s core product businesses — PC chips, server processors, networking silicon — must stop losing ground to competitors.
That’s a lot of things that need to go right simultaneously.
But for the first time in a while, Intel has something it hasn’t had: momentum in the right direction, backed by names that command attention. Whether that momentum translates into durable competitive advantage or merely buys time is the multi-billion-dollar question hanging over the company’s next chapter.
Musk, characteristically, hasn’t said much publicly about the Intel partnerships. He doesn’t need to. The signal is the deal itself.
And for Intel, that signal — broadcast to customers, competitors, investors, and the U.S. government alike — might be the most valuable chip the company has produced in years.
WebProNews is an iEntry Publication