Anthropic secretly files IPO! Tracing through its sky-high compute contracts, is Neocloud’s hidden investment thread emerging?

The battle for the title of the first AI stock has officially begun.

In the early hours of today, Anthropic announced it has confidentially filed an S-1 registration statement with the U.S. Securities and Exchange Commission (SEC). The company, founded just five years ago, last week completed a $65 billion Series H funding round at a staggering $965 billion valuation—historically surpassing its longtime rival OpenAI in the capital markets.

Source: Anthropic

However, powering this 'near-trillion-dollar' behemoth is an insatiable appetite for computing power. To meet the explosive growth in model training and inference demands, Anthropic has recently been aggressively acquiring foundational infrastructure. Most notably, last month it struck a landmark agreement with SpaceX: Anthropic will fully utilize SpaceX’s Colossus 1 data center in Memphis, Tennessee, under a five-year contract (through May 2029), paying $1.25 billion per month.

For Anthropic, this is a necessary capital expenditure to maintain its technological edge; but for market participants, this invoice serves as a precise 'treasure map.'

Looking through this deal, we not only see exceptionally high cash conversion efficiency from underlying computing infrastructure but also detect a highly explosive, trade-oriented hidden trend:The full-scale rise of Neocloud (next-generation cloud providers).

I. The Underlying Ledger of Computing Infrastructure: High-Certainty Cash Flow Returns

Just how profitable is this $1.25 billion-per-month infrastructure business? ARK’s chief futurist has done the math, clearly revealing the astonishing financial returns behind compute leasing:

Stunning revenue conversion: Anthropic pays SpaceXai up to $24 billion in revenue per gigawatt (GW) of nameplate capacity.

Extreme leverage effect: SpaceXai’s construction cost for these facilities is approximately $29 billion per GW. If this contract runs its full five-year term as expected, cumulative pre-tax cash flow will surpass $50 billion.

Potential upside: The above calculation is based only on the conservative assumption that SpaceX leases out its GB200/300 compute clusters. If the mix shifts to leasing all of Colossus I (primarily H100/H200) plus a portion of Colossus II (primarily GB200/GB300), total cash flow over five years could even surge to $60 billion.

Source: ARK

The core logic behind these figures is this: In a world suddenly realizing it is critically short on compute assets, the ability to convert cash into physical compute infrastructure with exceptional efficiency—and lock in long-term contracts—places a company at the strategic and profit sweet spot.

2. Spillover Effect of Compute Power: Traditional Cloud Exits, Neocloud Rises

As major tech firms’ compute capacity arms race aggressively spills outward, market capital is shifting its focus. Historically, when discussing cloud services, the default choices were always Amazon AWS, Microsoft Azure, and Google Cloud. However, facing exponentially surging demand, the architectures of these traditional cloud giants are now encountering dual constraints:

1. Severely constrained capacity: Traditional cloud giants act as both referee and player—their own AI products (such as Gemini and Copilot) are voraciously consuming internal compute resources, leaving them unable to allocate sufficiently large and dedicated compute pools to meet external large-model companies’ demands.

2. Antitrust concerns and vendor lock-in risks: Leading players like Anthropic or OpenAI must seek neutral, diversified compute infrastructure to break free from compute hegemony and reliance on a single supplier.

This pronounced 'compute spillover effect' has directly thrust Neocloud into the spotlight of the historical stage. Unburdened by the bloated legacy general-purpose businesses of traditional giants, Neocloud providers leverage exceptional agility and aggressive capital expenditure to aggressively acquire GPUs across the network and rapidly convert them into pure, high-performance AI compute pools.

Today, new players like $CoreWeave (CRWV.US)$ 、 $NEBIUS (NBIS.US)$ 、 $IREN Ltd (IREN.US)$ are quietly becoming the most efficient and indispensable 'top-tier arms dealers' of the large-model era. And in the AI gold rush, those selling shovels and weapons are often the first to hear the sound of coins clinking.

3. Core Trading Theme: How to Position Yourself in the Compute Leasing Sector?

previouslySpaceX Exposes High-Priced Compute Contracts, Nebius Announces GPU Price Hikes… Has the 'Turning Point' Arrived for U.S. Stock Compute Leasing Sector?A previous article梳理ed the relevant industry chain, as detailed below:

1. New Cloud Service Providers: The 'Pure Water Sellers' of the AI Era

These companies are a new breed born alongside the demand for large model training, focusing exclusively on offering high-end GPU computing power leasing services. They are typically deeply integrated with hardware giants like NVIDIA.

$CoreWeave (CRWV.US)$：Currently, the purest and most aggressively expanding AI-dedicated cloud service provider in the market—and NVIDIA’s 'favorite child,' having received direct investment and priority hardware allocation from NVIDIA. Thanks to extremely high GPU density and network architectures optimized specifically for AI workloads, it delivers more efficient and cost-effective per-unit computing power than traditional cloud providers. The market is closely watching its capacity expansion pace and NVIDIA’s future stakeholding moves.

$NEBIUS (NBIS.US)$：A rising computing power player with European technical roots, which has also recently secured investment from NVIDIA. It has drawn significant attention for announcing GPU price hikes and making massive investments to lock in Bloom Energy fuel cells—in other words, it was the first to break through the 'power shortage' bottleneck by building its own independent power generation network to secure computing capacity. Such Neoclouds with forward-looking energy strategies hold the strongest pricing power in this seller’s market marked by supply-demand imbalances.

$IREN Ltd (IREN.US)$: Successfully transitioned from Bitcoin mining to a 100% renewable-energy-powered AI computing center.Its advantage lies in readily available green power reserves and existing data center infrastructure, enabling it to rapidly deploy NVIDIA’s latest architectures.

Notably, Jensen Huang officially announced NVIDIA’s new DSX platform yesterday, with the above three companies listed among its partners. CoreWeave stated it has completed the industry’s first activation and validation of NVIDIA’s Vera Rubin NVL72 platform.

Fellow investors who are interested can click to view more.Breaking Down Jensen Huang’s Taipei Speech: From NVIDIA’s DSX Compute Ecosystem to AI PC 2.0—Which Companies Deserve Close Attention?~

$WhiteFiber (WYFI.US)$：An emerging infrastructure provider specializing in high-bandwidth fiber networks and interconnectivity between computing nodes. In hyperscale clusters (on the order of 100,000 GPUs), data transmission latency directly impacts computing efficiency. These companies solve critical pain points in intra- and inter-data-center connectivity.

2. High-Performance Computing (HPC): The 'Energy Landlords' Controlling Foundational Infrastructure

This segment includes numerous former cryptocurrency miners. In today’s AI computing crunch, their 'compliant grid capacity' and 'cooling infrastructure' have become invaluable assets coveted by tech giants.

$Applied Digital (APLD.US)$：is a next-generation data center developer specifically designed for high-performance computing (HPC). It has just signed a massive-scale AI factory lease agreement, upgrading its business model from 'buying GPUs and renting them out' to 'directly collecting super rents from tech giants.' Its core moat lies in its ability to secure land and power supply, as well as its engineering and delivery capabilities for building liquid-cooled data centers.

$TeraWulf (WULF.US)$& $Core Scientific (CORZ.US)$ ：are HPC transformation pioneers with substantial power reserves. WULF focuses on zero-carbon/nuclear-powered electricity, while CORZ leverages its extensive infrastructure footprint to frequently secure long-term AI hosting contracts. Specifically, hyperscalers are rushing to purchase power allocations—often at a premium—from these HPC companies that offer clean energy and ready-to-use facilities, reflecting a pure 'monetizing power' logic.

$MARA Holdings (MARA.US)$, $Riot Platforms (RIOT.US)$, $CleanSpark (CLSK.US)$, $Hut 8 (HUT.US)$Wait for:is among North America’s leading cryptocurrency mining giants, gradually diversifying part of its high-energy-consuming assets into AI-focused high-performance computing (HPC) businesses. As traditional 'energy arbitrageurs,' the market is now revaluing the real option value embedded in their hundreds of megawatts of available power capacity.

IV. Conclusion: Seeking Infrastructure Plays Amid the 'Gold Rush,' While Remaining Cautious of High Volatility Risks

Anthropic’s filed S-1 prospectus and its exorbitant compute bills essentially represent an early move to secure future AI compute pricing power. This race to become the 'first AI stock' may appear on the surface to be a showdown between large model technologies, but at its core, it is driving a full-blown infrastructure boom and a supercycle of capital expenditure.

This provides the market with a relatively clear lens through which to assess near-term earnings visibility: from Neoclouds holding pure compute resources, to fiber-optic network providers precisely addressing data center interconnect (DCI) pain points, to HPC firms controlling foundational energy and liquid cooling infrastructure—a complete infrastructure thread spanning compute, networking, and power has now emerged.

However, high upside often comes hand-in-hand with high volatility. While focusing on structural opportunities in the compute leasing segment, investors should also closely monitor the following potential risks:

Liquidity Risk under the Asset-Heavy Model: Neocloud commonly adopts a capital-intensive, asset-heavy expansion model with high capital expenditures (Capex), making it highly sensitive to changes in the macro interest rate environment and financing costs.

Hardware Delivery and Supply Chain Delays: Capacity expansion of new computing power centers heavily depends on timely delivery of upstream chips. Any bottlenecks in NVIDIA’s next-generation architectures (e.g., Blackwell, Rubin, etc.) or in underlying optical communication components would directly delay computing power providers’ capacity ramp-up and cash flow conversion timelines.

Backlash from Slower-than-Expected Commercialization on the Application Side: The current 'seller’s market' for computing power is built upon large tech firms’ cost-insensitive arms race. If commercialization of downstream large language models slows or inference-side demand fails to materialize as expected, the lofty valuations of computing infrastructure will face repricing pressure.

Amid this epic industry trend, the core logic lies in earnings realization and policy alignment. Maintaining sharp awareness of capital flows across the supply chain, while retaining respect for market volatility, may be the key to navigating this AI cycle sustainably.

546K Views