Decoding SoftBank’s Intel gamble: Son’s bold move to connect ARM and chip manufacturing

As SoftBank unveiled a $2 billion Intel investment Monday, the Financial Times reported that prior talks between founder Masayoshi Son and Intel CEO Lip-Bu Tan explored ways to back Team Blue’s chipmaking business, including a potential outright buy of its foundry division—an option that ultimately fell through.

The investment in the struggling U.S. chip giant has stirred wide curiosity. At $23 a share—less than half the chipmaker’s price at the end of 2023—the entry looks like a bargain, as Nikkei’s Chinese edition points out. But, the scale of the bet, coupled with Intel’s current struggles, suggests this could be more than a financial play. What strategic considerations might be driving Masayoshi Son’s bold move?

Intel paradox: Son’s surprising manufacturing bet
As Nikkei’s Takashi Sugimoto notes, Son has rarely backed manufacturers in the tech industry. His only major attempt was SoftBank’s 1996 purchase of an 80% stake in Kingston Technology for $1.5 billion (¥160 billion). However, the deal strained SoftBank’s finances and was sold just three years later. As per Nikkei, Son himself later conceded it was a failure, remarking, “I’m not interested in manufacturing.”

In addition to concerns about SoftBank’s rare bet on manufacturing, the 2% stake also deviates from Son’s typical strategy of securing large ownership positions for strategic influence, raising questions about Intel’s potential integration within SoftBank’s portfolio, as Nikkei notes.

Meanwhile, as Son rides the AI wave with SoftBank’s Stargate Project, Nikkei highlights Intel’s innovation gap: once unrivaled in PC CPUs, the company ceded mobile to Qualcomm, and now trails far behind NVIDIA in AI chips. This creates an apparent mismatch between Son’s AI-focused strategy and Intel’s current market position.

Hidden agenda
According to Investopedia, citing analysts, the move could signal SoftBank’s chip ambitions, as the company may be gearing up to build Arm CPU tiles, AI XPUs, and Ampere chips tied to the Stargate Project. Meanwhile, partnering with Intel to manufacture these chips could also help Son deliver on his $100 billion U.S. investment pledge, as noted by the report.

Meanwhile, Nikkei explains that chips are seen as the key weapons of the AI era, and Masayoshi Son’s strategy actually began nine years ago with the 2016 acquisition of UK-based ARM for $32 billion. As the report notes, ARM, a leader in semiconductor design and the “brain” of smartphones, became Son’s springboard into AI.

Arm’s ambitions to develop its own chips have been circulating for some time. Back in late July, Reuters reported that the company was increasing investment in potential chiplet and integrated solution development. As Financial Times previously reported, Arm CEO Rene Haas is expected to unveil the company’s first in-house chip as early as this summer, with Meta reportedly lined up as its debut customer.

In this context, Intel’s strategic value becomes clearer. As highlighted by Nikkei, ARM focuses solely on chip design and lacks manufacturing facilities, while Intel operates foundry services but faces underutilized capacity.

This creates a natural partnership opportunity: Intel can provide ARM with advanced manufacturing capabilities, while ARM offers Intel access to the growing mobile and IoT markets where power efficiency is crucial—a mutually beneficial collaboration, Nikkei adds.

Source: TrendForce, Taiwan