Scale, Scarcity, Strategy: Three Faces of Technological Power

Key Takeaways

• Oracle’s record-breaking $332 billion in bookings signals a paradigm shift from cloud vendor to geopolitical GPU supplier, redefining its role in the AI era.
• Synopsys quietly commands global leverage through its dominance in EDA software, a chokepoint so strategic it was traded for rare earths in a U.S.-China standoff.
• Intel’s partial nationalization by the U.S. government transforms it into a sovereign utility, blurring the lines between corporate performance and national defense strategy.

Oracle: Transformation at Scale

Every so often, a corporate earnings report delivers a number so overwhelming that it erases all the usual debates about revenue guidance, margin compression or quarterly beats relative to analyst expectations. For Oracle, that number was $332 billion in bookings in a single quarter.¹

This was not incremental growth. It was the largest bookings number ever reported in enterprise software, so large that it transformed how the market and policy makers must view Oracle. With contracts stretching up to a decade, including one annualizing at $30 billion, Oracle was no longer just a software vendor.² It had become a graphics processing unit (GPU) data center operator of geopolitical consequence.

Ironically, Oracle missed on several traditional measures: revenue growth underwhelmed, cloud infrastructure growth lagged peers, and earnings per share (EPS) fell below consensus. But the bookings overwhelmed all else.³ The narrative had shifted. Oracle was no longer defined by quarterly comparisons to Amazon Web Services (AWS) or Azure. It was defining a new industrial category altogether.

This is theater at its highest form: the ability to redirect the entire conversation with a single figure. Oracle has seized control of the compute narrative, positioning itself as custodian of the most valuable resource of the AI era, GPU capacity.

The Stargate Project as Prelude

The company’s Stargate project,⁴ once seen as a curiosity, now looks like a rehearsal. By collaborating on mega-scale data centers, Oracle tested its ability to think like an infrastructure operator rather than a software vendor. The experiment prepared it for this moment: where compute capacity is not merely a product, but a strategic commodity.

Oracle is now in the business of signing contracts that resemble long-term oil supply agreements. Its capital expenditure (capex), projected at $35 billion for FY26,⁵ is staggering. More than half of its annual revenue will be reinvested into GPU infrastructure. But these outlays buy more than machines; they buy relevance. They buy durability. They buy geopolitical positioning.

In that sense, Oracle is less a hyperscaler competitor than a strategic supplier. Like Saudi Aramco has done with oil, Oracle is building the pipelines through which nations and corporations will access compute.

Synopsys: The Silence of Indispensability

While Oracle created a spectacle with its blockbuster results, Synopsys exerts its influence in near silence. Its recent earnings told a more challenging story: sales of $1.74 billion came in below guidance, design intellectual property (IP) revenue was sharply weaker at $428 million versus the $552 million that analysts expected, and EPS fell short. Analysts also had concerns about pro forma noise from the Ansys merger.⁶

But none of this undermines Synopsys’s strategic position. That’s because Synopsys owns a large chunk of one of the most irreplaceable bottlenecks in the global economy: electronic design automation (EDA) software.

EDA is the invisible scaffolding of the semiconductor world. It is the set of tools without which chips cannot be designed, tested or manufactured. China can manufacture central processing units (CPUs), GPUs and memory. It can scale fabs. It can write software. But it cannot yet replace EDA. That space remains effectively controlled by three companies—Synopsys, Cadence and Siemens—that collectively own nearly 80% of the EDA market.⁷

This concentration creates a quiet form of power. Unlike GPUs or fabs, EDA doesn’t appear on earnings slides or consumer products. But without it, the semiconductor ecosystem would grind to a halt.

Minerals for Math: The July 2025 Swap

This hidden power became visible in July 2025. Facing a standoff over tariffs and supply chain disruptions, Washington and Beijing struck a deal: the U.S. lifted restrictions on EDA software exports in exchange for China resuming rare earth shipments.⁸

The symbolism was striking. The U.S. traded access to software tools for access to physical minerals. Commodities for code. Minerals for math.

Each side weaponized its bottleneck. China cut off rare earths, vital inputs for electronics and defense. The U.S. restricted EDA, vital inputs for chip design. Both sides blinked, restoring flows. But the precedent was set: EDA is as strategically vital as rare earths.

Unlike minerals, however, EDA cannot simply be mined. It represents decades of accumulated complexity and institutional knowledge. It is not a natural resource but a cognitive resource. And that makes Synopsys’s power even harder to replicate.

Intel: Too Important to Fail

There is a third face of technological power, one that neither Oracle’s theater nor Synopsys’s silence fully captures: direct state intervention. That role is illustrated by Intel.

When the U.S. government took a roughly 10% equity stake in Intel, it was more than industrial support. It was partially nationalizing the semiconductor supply chain.⁹

This stake reframed how investors must view policy. Grants can be withdrawn; credits can be revised. But equity implies alignment of destinies. By owning a share of Intel, the U.S. government has committed not just to supporting the company, but to ensuring its survival, relevance and strategic role in national defense.

Intel thus becomes not only a company, but a policy instrument. Its balance sheet is intertwined with Washington’s geopolitical agenda. Its research and development (R&D) roadmap is no longer just a question of market share against Advanced Micro Devices (AMD) or Taiwan Semiconductor Manufacturing Co. (TSMC); it is a matter of national capacity.

Three Archetypes of Power

Put together, Oracle, Synopsys and Intel represent three archetypes of technological power in the 2020s:

• Oracle – Visible Scale. Spectacular contracts that reshape narratives and reposition the company as a custodian of compute.
• Synopsys – Invisible Indispensability. Quiet control over an irreplaceable chokepoint, valuable enough to be traded for rare earth minerals.
• Intel – Sovereign Alignment. Direct state involvement, with government equity ensuring survival and signaling strategic intent.

This trinity covers the spectrum of how geopolitical power is manifesting in the AI competition.

For Investors: Contrasting Risk Profiles

For investors, these distinctions matter.

Oracle’s model is bold but capital-intensive. It is betting on a future where GPUs remain scarce and demand for compute only accelerates. If that thesis holds, its revenue base will compound. If not, the heavy capex may weigh.

Synopsys is subtler. It has faced earnings volatility, but its moat is structural. No advanced chips are made without EDA. Export controls may tighten, but as the July trade showed, the U.S. ultimately views Synopsys as too valuable to restrict for long. It is less a growth story than a perpetual rent story, deeply embedded in the semiconductor lifecycle.

Intel offers a different set of risks entirely. With the U.S. government as a shareholder, downside could be at least partially cushioned. But upside is capped by the reality that Intel is now a strategic utility. Investors must recognize that returns will increasingly be tied to national policy goals, not just market competition.

The Geopolitical Thread

What unites all three is geopolitics. Oracle’s contracts make sense only in a world where nations hoard compute as a strategic resource. Synopsys’s indispensability only becomes visible in the context of trade wars. Intel’s stake only exists because semiconductors are viewed as sovereignty-critical.

The lesson is clear: In the AI era, the boundary between corporate strategy and state policy has dissolved. Companies are no longer merely competing in markets; they are instruments of national power.

Conclusion: The Triangle of Power

In the 20th century, power revolved around oil, steel and finance. In the 21st, it revolves around compute, certain design tools and semiconductors. Oracle has dazzled with its recently reported quarterly bookings. Synopsys disappears into indispensability, bartered like a rare earth mineral. Intel binds itself to the U.S. state, blurring the line between corporation and government.

For investors, this is both an opportunity and a warning. The rewards of scale, scarcity and sovereignty could be immense. But the risks are equally real. Policy is not a tail risk anymore; it is the operating environment.

The real question, then, is not which of these companies is best positioned for the next quarter. It is which form of power—visible scale, invisible indispensability or sovereign alignment—will prove most durable in a world where technology is not just an industry but the frontline of geopolitics.

Originally posted on September 25, 2025 on WisdomTree blog

FOOTNOTES AND SOURCES

DISCLOSURES