1. How I First Discovered NVIDIA
My journey with NVIDIA began in 2016, right at the onset of the crypto mining boom. But unlike the crowd, I wasn’t drawn in by the buzz around mining rigs. I had a far longer view.
As a television engineer who played a key role in Australia’s digital broadcast revolution in the early 2000s, I had firsthand experience with the architecture of digital pipelines. I knew the role GPUs played in encoding, decoding, and managing video data—and how that would expand beyond pixels to data across every industry.
Later, my work in autonomous mining robotics only deepened that conviction. I came to believe that any system trying to mimic human perception and cognition—whether for self-driving cars or industrial automation—would need massive, real-time parallel compute power. That wasn’t a CPU job. It was—and is—a GPU-native world.
2. The Spark: A Broader Vision for GPUs
When I first looked closely at NVIDIA, I saw a company not just making chips—but building foundational infrastructure for the digital age. Their GPUs were becoming the nervous system of the world’s most demanding systems: graphics, simulation, autonomous vehicles, scientific computing, and eventually AI.
That understanding passed what I call my “systems sniff test.” I wasn’t reacting to trends—I was watching the evolution of digital value chains and betting on the enabler.
3. Investigative Process
Over four years, I dollar-cost averaged into NVIDIA—starting with a small position and letting my research build conviction.
I read every earnings call. I studied CEO Jensen Huang’s public commentary. I dug into product roadmaps—especially around CUDA, the NVIDIA Drive platform, and later, their GPU-accelerated data centre stack. I tracked their acquisitions (like Mellanox), the developer ecosystem, and their growing moat in developer mindshare.
What impressed me most was their visionary execution—not just technology leadership, but architecture control, vertical integration, and bold strategic pivots years ahead of competitors.
4. My Original Thesis
My initial thesis was simple: GPUs are the new engines of intelligence.
I believed NVIDIA would dominate in:
- Digital content acceleration (video, gaming, AR/VR),
- Self-driving vehicles, where real-time decision-making at the edge is critical, and
- High-performance computing across scientific, financial, and defense sectors.
They were building tools for industries I had worked in—and more importantly, for industries that didn’t exist yet.
5. Role in My Portfolio
I began with a modest starter position—as always, testing both market and personal conviction. Over the years, that grew to 3% of capital as I scaled in with conviction.
Today, NVIDIA represents 12% of my portfolio, making it my largest position. This was not a speculative moonshot—it was an earned, compounding conviction over a 9-year journey.
It lives firmly in my G1 Fund – Core Compounders, but also expresses themes in my G2 and G3 funds due to its central role in AI, robotics, and infrastructure computing.
6. The Leadership Team
Jensen Huang is, to me, one of the great architect-CEOs of our time—alongside the likes of Jobs, Nadella, and Bezos. He blends technical mastery with missionary zeal. His leadership isn’t about quarterly results—it’s about 20-year arcs.
Huang understands that developers are the most valuable assets in modern compute ecosystems—and NVIDIA has relentlessly cultivated that base.
His clarity of thinking, charisma, and precision of execution have created a culture of first principles thinking and iteration—rare in public companies of this scale.
7. Experience as a Public Company Leader
NVIDIA’s capital allocation record is exemplary: investing early in trends that others ignored (AI in 2015, self-driving in 2016, data centre integration in 2019), and staying focused when markets were distracted.
Their operational leverage, R&D velocity, and discipline in ecosystem development have made them the default choice in high-performance compute.
They’ve shown resilience—weathering the 2018 crypto crash, chip cycles, and pandemic shocks—without losing strategic focus.
8. Performance Against Key Metrics
NVIDIA consistently outperforms on the metrics I care most about:
- Gross Margins well above 60%—demonstrating pricing power and architecture advantage.
- Rule of 40: Their growth and profitability metrics together routinely exceed 60–70, even amid macro headwinds.
- Return on Invested Capital (ROIC) has been among the highest in the industry—clear evidence of compounding efficiency.
More importantly, they lead in non-financial KPIs: developer adoption, ecosystem stickiness, and enterprise standardization.
9. What I’m Watching for Now
My current thesis is centred around the Data Centre Explosion—driven by the rise of:
- Large Language Models,
- AI training and inference at scale,
- Edge compute for robotics and real-time systems,
- And AI-as-a-service platforms being built by hyperscalers and enterprise clients.
I believe NVIDIA is not just a chip company. It’s a platform provider for the intelligence economy.
I’m watching:
- Supply chain expansion: H100, B100, and custom silicon.
- Software adoption: CUDA, Triton, Omniverse, and AI Foundation Models.
- Regulatory risk in China and the U.S.
- Entry of new competitors (Intel, AMD, custom ASICs).
NVIDIA’s Business Verticals: TAM, Revenue, and Market Share
1. Data Center & AI Infrastructure
- Total Addressable Market (TAM): Projected to reach $1 trillion by 2028, driven by the proliferation of AI workloads, large language models (LLMs), and accelerated computing demands .
- NVIDIA’s FY25 Revenue: Approximately $115 billion from the data center segment .
- Market Share: Dominant position with 98% of data center GPU shipments in 2023 .
2. Gaming (Discrete GPUs)
- TAM: Estimated at $50–60 billion annually, encompassing PC gaming, cloud gaming, and eSports.
- NVIDIA’s FY25 Revenue: Approximately $9.1 billion from the gaming segment .
- Market Share: Leading with 90% of the discrete GPU market as of Q3 2024 .
3. Automotive (AI, Robotics, Autonomous Driving)
- TAM: Projected at $300 billion, encompassing autonomous driving, AI cockpit systems, and vehicle-to-cloud services .
- NVIDIA’s FY25 Revenue: Approximately $1.7 billion, reflecting a 5% increase from the prior year .
- Market Share: Significant presence in the automotive hypervisor industry, contributing to a combined 58.3% market share alongside key partners .
4. Professional Visualization (Workstations, AR/VR, Industrial Design)
- TAM: Estimated at $20–30 billion, covering sectors like architecture, engineering, media, and entertainment.
- NVIDIA’s FY25 Revenue: Approximately $1.88 billion from the professional visualization segment .
- Market Share: Strong position in the professional GPU market, though specific percentage figures are not publicly disclosed.
10. What Would Break My Thesis
- A complete breakdown in developer loyalty or ecosystem strength (unlikely).
- Structural regulatory barriers limiting market access or export.
- A Moore’s Law–level shift away from GPU-dominant architectures (e.g., photonics, quantum).
- Jensen Huang stepping down without a visionary successor.
But even with these risks, I see NVIDIA continuing to benefit from Mach 2-scale data centre demand, where compute scales exponentially and latency becomes currency.
Final Reflection
This isn’t just my best-performing stock—it’s the embodiment of my investing philosophy: long-term, systems-driven, asymmetric in potential, and grounded in lived industry insight.
If you’re reading this decades from now, and NVIDIA is a foundational layer of every intelligent system on Earth, you’ll know why I invested early, held through the noise, and remained steadfast.
I didn’t just bet on chips. I bet on the infrastructure of the future
