Can Intel become SoftBank’s next Arm? Why the $2bn bet could reshape its AI hardware strategy

SoftBank Group Corp. (Tokyo: 9984) made waves in August 2025 when it disclosed a $2 billion equity investment in Intel Corporation (NASDAQ: INTC), acquiring shares at $23 apiece and securing a roughly 2 percent ownership stake. While the transaction was initially framed as a financial vote of confidence, the move has since triggered deeper analysis across markets, drawing comparisons to SoftBank Group’s earlier, high-conviction bet on Arm Holdings nearly a decade ago.

Three months later, as institutional investors digest SoftBank Group’s Q3 2025 reshuffle and Intel Corporation begins to reassert its foundry ambitions, the deal is being interpreted less as a passive portfolio addition and more as a signal of strategic intent. In shifting capital away from mature growth plays like NVIDIA Corporation and Oracle Corporation, SoftBank Group appears to be reorganizing its AI thesis around infrastructure-scale bets rather than model-centric narratives.

The investment comes as SoftBank Group founder Masayoshi Son intensifies his focus on building the physical backbone of artificial intelligence. For some observers, Intel Corporation may now represent the hardware complement to what Arm Holdings once offered in chip design, a foundational pillar in a new AI stack. Whether the parallel holds will depend on execution, scale, and how Intel Corporation’s manufacturing resurgence aligns with SoftBank Group’s broader infrastructure agenda.

Does SoftBank Group see Intel Corporation as the next Arm Holdings—only deeper in the stack?

When SoftBank Group acquired Arm Holdings in 2016, it aimed to dominate the chip design layer of computing, securing licensing revenue from a wide ecosystem of mobile and embedded processors. The strategy worked well in the pre-AI era, where controlling intellectual property and ecosystem scale offered significant financial leverage without the burden of capital-intensive fabrication.

By contrast, the Intel Corporation investment shifts that focus downward into manufacturing. SoftBank Group’s August 2025 statement made clear that the move was based on its conviction that advanced semiconductor manufacturing and supply would expand significantly in the United States, with Intel Corporation positioned to lead that shift. In essence, if Arm Holdings was the blueprint, Intel Corporation could become the factory.

This thematic alignment is playing out in the backdrop of geopolitics, reshoring mandates, and the AI compute race. Intel Corporation remains one of the few American firms with the capability and ambition to scale advanced-node foundry services. SoftBank Group’s $2 billion injection gives it a stake in that turnaround, but also places it closer to the substrate of AI infrastructure.

Although the stake does not include a board seat or voting influence, industry insiders view the investment as more than just a financial maneuver. Given SoftBank Group’s recent push into AI data centers through projects like Stargate, and its ongoing exposure to chip design via Arm Holdings, this exposure to Intel Corporation’s fabrication capabilities effectively rounds out its vertical position in the AI stack.

What progress has Intel Corporation made since SoftBank Group’s entry—and what must it deliver next?

Since SoftBank Group’s investment in August 2025, Intel Corporation has taken cautious but visible steps to validate its long-term foundry thesis. In its third-quarter earnings reported in late October, the company posted revenue of $13.7 billion, up 3 percent year-over-year. While the client computing group showed signs of stability, with growth of around 5 percent, data center and AI revenue dipped 1 percent and foundry revenue fell 2 percent, highlighting the complexity of Intel Corporation’s turnaround.

Execution remains a key concern. The company must ramp up its foundry services to win external customers, scale advanced nodes like Intel 18A and Intel 14A, and restore trust among hyperscalers and chip startups looking for sovereign fabrication capacity. CEO Lip-Bu Tan, appointed earlier this year, has promised aggressive transformation, but markets remain watchful.

What makes this more than a typical recovery story is the macro environment Intel Corporation is operating in. With U.S. government interest in taking a direct stake in Intel Corporation, and federal incentives from the CHIPS and Science Act already in play, the company is embedded in a national narrative around industrial resilience and AI hardware independence. SoftBank Group’s move coincides with this momentum, but does not mitigate the execution burden that lies ahead.

SoftBank Group’s thesis likely goes beyond quarterly metrics. Its broader infrastructure strategy, spanning Arm Holdings, Ampere Computing, Stargate, and now Intel Corporation, appears to be positioning for end-to-end AI enablement, from chip design to deployment. However, Intel Corporation’s capex-heavy model and historical underperformance still represent real headwinds.

How does this compare to the Arm Holdings play and what risks make this bet more complex?

Unlike Arm Holdings, which delivered margin-rich licensing growth with minimal overhead, Intel Corporation is a vertically integrated enterprise with heavy operating leverage. SoftBank Group’s earlier Arm Holdings thesis benefited from timing: mobile growth was surging, Arm Holdings’ business model was structurally efficient, and ecosystem lock-in ensured high monetization per design win.

Intel Corporation, on the other hand, is a turnaround story that needs to rebuild its reputation across multiple verticals, which include foundry services, AI acceleration, and packaging innovation. Taiwan Semiconductor Manufacturing Company and Advanced Micro Devices have eroded Intel Corporation’s market share across these domains. Reclaiming lost ground will require not just capital but clear technical wins.

SoftBank Group’s $2 billion investment, while significant, represents a minority holding that does not provide operational control or direct influence over product roadmaps. If Intel Corporation’s recovery falters or foundry utilization fails to scale, the financial upside could be limited. The market’s patience will be tied to signs of external customer adoption, cost discipline, and clarity around Intel Corporation’s ability to compete with established foundry giants.

There is also the geopolitical dimension. Intel Corporation’s alignment with U.S. policy means it may attract increased government scrutiny and may be subject to export controls, allocation mandates, or defense-sector obligations. SoftBank Group is entering this ecosystem not as an operator, but as a long-term thematic investor whose return profile is now linked to American industrial strategy and fabrication nationalism.

What does this say about SoftBank Group’s evolving AI infrastructure thesis?

Over the past year, SoftBank Group has been moving away from model-centric AI bets and toward physical enablers of the AI economy. The August 2025 Intel Corporation investment must be viewed alongside its exits from NVIDIA Corporation and Oracle Corporation and the capital reallocation into data centers and embedded finance platforms.

This repositioning suggests that SoftBank Group is betting on scarcity and scale, which are the two things that are becoming most valuable in the AI infrastructure cycle. Chip design IP, foundry capacity, and power-dense AI compute facilities are the ingredients SoftBank Group appears to be assembling under a new investment doctrine. Intel Corporation is the manufacturing node in this vision.

By late 2025, as AI model proliferation begins to saturate and the bottlenecks shift to fabrication and supply chains, the value in companies like Intel Corporation could re-rate. However, this outcome is not assured. Execution timelines are long, risks are high, and SoftBank Group’s stake, though large, lacks direct governance tools.

Still, this remains one of the more strategically aligned moves SoftBank Group has made in recent memory. If Intel Corporation’s turnaround holds and its foundry services become central to AI sovereignty narratives, SoftBank Group may once again find itself at the center of a foundational technology cycle, this time not through design licenses, but through the machines that build the future of compute.