Nicolas Sauvage: The Maverick Investing in AI’s Invisible Backbone

Nicolas Sauvage believes it takes four years for the best bets to look obvious — thinking that he shared on stage last week at StrictlyVC’s San Francisco event, which TDK Ventures co-hosted.

**Key Takeaways**

* **The “Four-Year Rule”:** Nicolas Sauvage, founder of TDK Ventures, operates on the principle that truly impactful investments only become “obvious” four years after the initial bet, requiring deep foresight into future technological bottlenecks.
* **Strategic Corporate Venturing:** TDK Ventures was founded on a unique mandate to identify the next big opportunities for TDK and anticipate potential industry disruptions, moving beyond traditional corporate VC models.
* **Betting on Unseen Bottlenecks:** From AI inference chips like Groq to physical AI with specific utility and the resurgence of CPUs for AI orchestration, Sauvage consistently targets overlooked infrastructure and dexterity challenges that will define the next wave of technological evolution.

—

## The “Four-Year Rule”: Unveiling Tomorrow’s Obvious Bets Today

Nicolas Sauvage, the visionary behind TDK Ventures, isn’t just investing in the future; he’s meticulously timing it. At last week’s StrictlyVC event in San Francisco, co-hosted by TDK Ventures, Sauvage articulated his core investment philosophy: the best bets only become undeniably obvious after a four-year incubation period. This isn’t just a theory; it’s the operational backbone of TDK Ventures, which he founded in 2019, now managing an impressive $500 million across four funds.

The prime example of this prescient thinking is AI chip startup Groq. Valued at $6.9 billion in its most recent funding round last fall, Groq was far from a mainstream bet when Sauvage first wrote a check in 2020. This was well before the generative AI explosion made infrastructure investments a clear-cut choice. Groq, founded by Jonathan Ross—one of the key engineers behind Google’s Tensor Processing Units—was laser-focused on inference: the demanding computational workload involved every time an AI model generates a response. Ross’s innovative approach involved building the compiler first, meticulously stripping down the chip architecture until, as Sauvage describes it, “you can’t remove one part and have it still work.”

While some might have viewed this as a niche pursuit, Sauvage, keenly aware of his parent company’s strategic needs, saw profound asymmetry. Unlike consumer hardware, which often faces inherent demand ceilings, the need for inference continually compounds with every new application and model. Though he couldn’t have predicted the exact explosion of demand driven by multi-agent AI systems, where a single query now branches into dozens of calls, Sauvage understood the underlying, inevitable growth trajectory.

## The Genesis of an Unlikely Venture: Forging TDK Ventures

The very existence of TDK Ventures is, by Sauvage’s own admission, an unlikely tale. A Japanese electronics conglomerate primarily known for magnetic tape might not seem like the most natural home for a cutting-edge venture fund. Sauvage, a Frenchman who joined TDK in Silicon Valley through an acquisition, lacked the conventional standing to propose such a radical idea to TDK’s Tokyo headquarters. “I’m not Japanese. I don’t speak Japanese; I don’t live in Tokyo,” he candidly shared with this editor.

Yet, driven by conviction, Sauvage persisted. Following two influential Stanford lectures—one championing the benefits of corporate VC, the other detailing its common pitfalls—he pitched his vision. His refusal to accept initial rejections eventually bore fruit, securing the green light to build a fund with a singular, strategic mandate: to uncover “What’s the next big thing for TDK, and what might kill it?” This guiding question ensures that TDK Ventures’ investments are not just financially sound, but also strategically aligned with TDK’s long-term future.

## Beyond AI Chips: A Portfolio of Future Bottlenecks

The portfolio Sauvage has meticulously assembled since 2019 reflects this deep understanding of impending bottlenecks. It’s replete with technologies that have only recently captured broader VC interest. This includes investments in solid-state grid transformers, sodium-ion batteries designed for data centers, and other alternative battery chemistries that strategically sidestep the geopolitical vulnerabilities inherent in lithium and cobalt supply chains.

The consistent discipline underpinning all these ventures remains the same: identify the critical bottlenecks four years down the line, then seek out the visionary founders who are already pioneering solutions.

## Peering into the Future: Sauvage’s Next Big Bets

The perennial question, of course, is what lies ahead. Sauvage is currently focused on several emerging areas he believes will define the next wave of innovation.

**Physical AI with Purpose:** He’s closely watching “physical AI,” specifically robots designed for highly specialized tasks, rather than general-purpose robotics. For instance, Agility Robotics, a company in his portfolio, addresses workforce shortages by focusing solely on the mundane but crucial task of moving objects in warehouses. Similarly, Swiss portfolio company ANYbotics develops ruggedized robots for hazardous environments, performing jobs where human presence is too dangerous. The common thread is clarity of purpose: these robots don’t attempt to do everything; instead, they reliably execute one demanding task.

**The CPU Renaissance for AI Orchestration:** Sauvage also foresees another significant shift in the compute stack. While GPUs have dominated AI training—the massive parallel computation required to teach models—and inference chips like Groq’s are redefining how those models “speak” with speed and scale, CPUs are poised for a resurgence. Though not the most powerful or fastest chips, their unparalleled flexibility makes them ideally suited for the branching, decision-making logic of AI orchestration. As AI agents increasingly delegate tasks, monitor progress, and navigate complex, multi-step choreographies, the humble CPU is emerging as the essential conductor.

**”Vibe Manufacturing” and the Dexterity Challenge:** Lastly, Sauvage is closely tracking a phenomenon he terms “vibe manufacturing,” illuminated by a recent report from venture firm Eclipse. This refers to the rapid, AI-assisted iteration of physical hardware prototyping, mirroring the “vibe coding” revolution in software. Chinese manufacturers, the report highlights, are dramatically compressing the design-build-test cycle for physical products in ways that Western supply chains are not yet equipped to match.

For Sauvage, this signals a crucial bottleneck. While AI models are advancing rapidly, making physical AI an inevitability, the missing piece is true physical fluency or “dexterity.” The nations and companies that master the ability to iterate on atoms as swiftly as others iterate on code will gain an insurmountable manufacturing advantage. This is the impending wave for which Nicolas Sauvage is strategically positioning TDK Ventures today.

—

**Bottom Line**

Nicolas Sauvage and TDK Ventures exemplify a rare breed of strategic corporate venturing. By meticulously identifying future bottlenecks four years in advance and fearlessly backing unconventional bets, Sauvage has cultivated a portfolio that not only drives significant financial returns but also strategically fortifies TDK’s future. His approach underscores that true innovation often lies in anticipating the unseen shifts in technology and infrastructure, rather than simply following current trends, solidifying TDK Ventures’ role as a key player in shaping tomorrow’s tech landscape.