Nvidia is evolving beyond a AI provider to become the "Central Bank" of a new "Computing Power" based economy, where computational power is akin to currency.
Through its strategic investments, such as in CoreWeave, Applied Digital, and Recursion Pharmaceuticals, Nvidia is not just supporting companies, but is architecting a new financial system where these companies act as "commercial banks" distributing and monetizing this computational currency.
This paradigm shift is changing how companies are valued, moving away from traditional metrics and towards the deployment and utilization of computational resources, indicating a fundamental change in tech capitalism and the future of AI development.
Building Computing Power as the New Currency
Nvidia's recently disclosed investment portfolio resembles a venture capital list, spanning cloud computing, data centers, AI biotech, edge computing, and low-power chips.
However, focusing solely on the investment targets risks missing a larger trend: Nvidia is laying the groundwork for a new financial order where computing power acts as a "quasi-currency," redefining tech capitalism.
Nvidia is effectively assuming the role of the "central bank" in this new financial system. If Nvidia is the central bank, its portfolio companies are the commercial banks. By allocating GPU computing power, Nvidia shapes the future tech economic landscape.
These companies aren't primarily inventing new technologies; they are repackaging raw computing power into new services, new supply chains, and new value propositions.
A Structural Shift: Computing Power as Asset and Currency
We stand at a pivotal moment of structural transformation: Computing power is not just an asset; it's becoming a new form of currency. As the deployment and training of AI models (like ChatGPT) become increasingly reliant on high-performance GPUs, the supply-demand value of these chips as tangible resources becomes evident.
Companies controlling computing resources will dictate the cost barriers for AI training. In other words, buying GPUs is no longer mere equipment procurement; it's akin to acquiring "computational liquidity rights" – a financial transaction. Nvidia supplies the GPUs; these companies take them, convert them into assets or cash flow, and reinvest to expand capacity. Isn't this "Modern Monetary Theory" (MMT) 2.0 for the tech-capital world?
Moreover, these companies aren't just competing with giants like Alphabet, Amazon, Apple, or Meta Platforms; they are forcing the "Magnificent Seven" to innovate relentlessly or risk obsolescence.
The Role of the Commercial Banks
Let's examine key Nvidia-backed companies and their roles in this computing power economy:
CoreWeave: This company secures thousands of Nvidia H100 and GB200 GPUs to build its own cloud platform. It doesn't just sell cloud services; it sells "AI-ready combat power." Unlike traditional Amazon or Google Cloud data centers, CoreWeave treats computing power as a liquid asset, redistributing it to enterprises needing immediate AI capability. Its valuation hinges not on EBITDA, but on GPU Deployment Rate – proving computing power is the core asset.
Applied Digital (APLD): APLD focuses on converting Nvidia's computing power into solutions outside Google's self-built data centers, repackaging it to offer new services and challenge the traditional data center model.
Weave Communications (WEAV): WEAV isn't a traditional hardware or SaaS company. Its foundation is computing power (data processing, analytics, model training, real-time response), enabling it to create entirely new business models, market demands, and value. It acts like a "loom," weaving together disparate endpoints, devices, data sources, users, and applications into a unified, intelligent network.
Recursion Pharmaceuticals (RXRX): RXRX leverages AI to apply computing power in biotech, challenging the inefficiencies of traditional big pharma and pioneering an AI-driven approach to drug discovery.
These companies disrupt traditional business logic.
Their core isn't product sales or profits; it's scaling computing power deployment and increasing turnover to generate value – much like commercial banks. They receive Nvidia's "currency" (GPUs) and transform it into assets, services, or cash flow.
Every new enterprise training AI models or leasing GPUs effectively allows these companies to "print money" by distributing and circulating computing power.
CoreWeave breaks cloud giants' monopolies by offering non-Amazon/Azure/Google Cloud access.
APLD provides non-Google data center options.
RXRX challenges inefficient traditional pharma with AI biotech.
Regardless of who ultimately wins among these players, Nvidia, as the central bank of computing power supply, remains the ultimate winner.
Implications for Investors
Investors relying solely on traditional PE ratios, EPS, or gross margins risk missing this structural shift towards computing power as an asset class.
These companies aren't merely growth stocks; they are evolving into financial agents ("AI Agents") for the computing power economy.
Imagine if the deployment of Saudi Arabia's GPT model or AI training demand becomes critically dependent on specific Nvidia GPUs. These chips gain intrinsic supply-demand value in the physical world. Companies controlling access to these computing resources will dictate the cost of AI training for enterprises.
This represents a fundamentally different capital logic – perhaps not just the next step for tech stocks, but another form of financial capitalism itself.
Nvidia isn't just investing in startups; it's investing in the early architecture of an entire "computing power monetary system." If the market embraces this valuation model, providers of deployed computing power could become rarer and more fundamental than SaaS companies or even cloud infrastructure.
The critical question for investors now: These companies won't just seek growth; they will aim to become agents of capital markets.
This is the future Jensen Huang is truly betting on.
Share this post