NVIDIA's 'AI Empire' Unveiled! Where Exactly Is Jensen Huang Investing His Money?

Through NVIDIA's capital moves, we may gain a clearer insight into the future evolution of the AI industry than from financial reports alone.

In this wave of AI advancements, $NVIDIA (NVDA.US)$ NVIDIA is undoubtedly a core beneficiary. However, beyond observing its revenue growth on financial reports, examining this giant's recent capital layout provides a key window into understanding the development trajectory of the AI industry over the next 3 to 5 years.

Over the past six months, NVIDIA’s strategic investment in listed companies has reached as high as $19 billion. This is not only to consolidate its existing industry alliances but also to use capital as leverage to actively catalyze the prosperity of downstream supply and demand ecosystems.

A review of NVIDIA's recent investment landscape reveals that its strategy has been precisely deployed infive core pillars. This is not only to consolidate its existing hardware advantages but also to build a powerful 'integrated hardware and software' AI ecosystem empire.

How is NVIDIA's investment empire being structured?

1. Strengthening foundational infrastructure: Breaking through the 'weakest link' in computing power and transmission

As AI cluster sizes scale up to hundreds of thousands of GPUs, bottlenecks in data transmission and underlying design are becoming increasingly apparent. NVIDIA’s investments aim to ensure extreme stability across the entire hardware supply chain while leapfrogging technological barriers.

Chip manufacturing and design： Investing in EDA software leader $Synopsys (SNPS.US)$ , aiming to utilize AI to assist in the design of next-generation GPUs, significantly shortening R&D cycles; entering $Intel (INTC.US)$ is about seeking strategic backup plans for advanced packaging and foundry capacity outside of Taiwan Semiconductor.

Optical communications and network interconnects: This is a highly predictable track in AI infrastructure. Strategically invest in optical communication giants $Lumentum (LITE.US)$ and $Coherent (COHR.US)$ to secure underlying capacity for 800G and future 1.6T CPO. At the same time, partner with leading network chip companies Marvell Technology and Enfabrica, which specializes in new I/O chips,to ensure efficient transmission of massive data between servers.

AI inference chips: Heavily bet on AI inference chip startups Groq, demonstrating early positioning in edge computing and specific high-speed inference scenarios to guard against potential technological disruptions.

II. GPU computing power cloud and data centers: Hold absolute authority in allocating computing resources

Traditional large cloud service providers (the Big Three cloud providers) are both NVIDIA's largest customers and active developers of proprietary AI chips, making them potential competitors. To mitigate the risk of being 'piped', NVIDIA is actively fostering its exclusive computing power ecosystem.

Pure GPU cloud provider: heavily invested in the leading pure GPU cloud company in North America $CoreWeave (CRWV.US)$, granting priority access to top-tier chips. At the same time, expanding into the European market through $NEBIUS (NBIS.US)$ and focusing on the developer market via NscaleandLambda Labs. The rise of these new computing power players will force traditional cloud giants to continue purchasing NVIDIA GPUs to maintain competitiveness.

Green computing infrastructure: Energy bottlenecks have become a widely acknowledged pain point for AI development. Investing in Crusoe (using flare gas from oil fields to power data centers) indicates that NVIDIA has begun to develop green energy solutions for data centers, suggesting that 'AI + Energy' will be a key focus for the market moving forward.

III. Frontier Technology: Exploring the Vast Potential of Computing Power

To maximize the value of computing power, NVIDIA is turning its attention to cutting-edge scientific fields that could bring about transformative changes in human society.

AI Drug Discovery: By leveraging GPU computing power to disrupt the traditional new drug development cycle, this is a highly scalable sector. NVIDIA has partnered with pharmaceutical giants $Eli Lilly and Co (LLY.US)$ and established $Recursion Pharmaceuticals (RXRX.US)$ as a benchmark for its BioNeMo platform, while also investing in chemical space exploration companies such asTerray, accelerating AI's implementation in biopharmaceuticals across the board.

Nuclear Fusion: Participated in investment in commercial nuclear fusion companiesCommonwealth Fusion Systems, further demonstrating its grand ambition to explore the ultimate clean energy for humanity and support the endless demand for AI computing power in the future.

4. Physical AI and Edge Computing: Cultivating the Next Generation of Computing Power Consumption Engines

As the training of large language models (LLMs) gradually converges, the enormous future demand for inference computing power will be supported by the physical world.

Telecommunications Infrastructure: Invest in telecom giants $Nokia Oyj (NOK.US)$ , aiming to push powerful AI inference capabilities to the edge of 5G/6G networks.

Humanoid robots: Physical AI is the next breakout point. Investing in leading humanoid robotics companiesFigure AI, indicating that future sensors, machine vision, and mechanical control technologies will experience massive demand.

Autonomous Driving:Investing in the rising star of end-to-end autonomous drivingWayve, promoting the application of pure data-driven AI models in real-world traffic scenarios.

5. Foundational large models and generative AI: Strengthening the CUDA software ecosystem moat

The leading advantage in hardware ultimately needs to be solidified through an impregnable software ecosystem.

Foundational large model giants: From participating inOpenAI's investments, to betting on Musk'sxAI, to open-source and enterprise-level model giantMistral AIandCohereNVIDIA, through its capital ties, ensures that the world's top large models continue to rely on the CUDA ecosystem for underlying training and inference optimization.

AI application layer and multimodal:Strategic positioning in AI search enginesPerplexityand leading AI video generation.RunwayParticularly, video generation and multimodal technologies demand far more computing power and memory bandwidth than pure text, which is the most favorable scenario for NVIDIA to generate high premiums from computational consumption.

Summary: Learning investment by following the giants' 'real investments'

In summary, NVIDIA's massive $190 billion capital deployment is not merely a financial investment but a meticulously planned 'ecosystem defense and expansion strategy.' From solidifying foundational hardware infrastructure, supporting affiliated cloud forces, to betting on physical AI, cutting-edge science, and multimodal large models, NVIDIA is using capital as leverage to preemptively lock in and catalyze enormous computing demands over the next 3 to 5 years.It has evolved from being a mere 'seller of picks' to becoming the 'kingmaker' controlling the lifeline of the AI industry.

For investors, this investment map provides an extremely forward-looking perspective."AI Industry Treasure Map"The future profit effect will no longer be limited to NVIDIA alone, but will ripple out along its strategic mainline, rapidly spreading to multiple细分sectors such as optical communications, green computing power, embodied intelligence, and AI pharmaceuticals.

In this epic supercycle of AI, understanding NVIDIA’s capital flow and following the巨头's significant investments to identify opportunities is the optimal solution for us to position ourselves in advance and accurately capture the next high-growth Alpha opportunity.

595K Views