Nebius, CoreWeave and the AI Infrastructure Boom: Why GPU Cloud, 2X ETFs and AI Buildout Economics Are Converging
The artificial-intelligence infrastructure trade is once again attracting significant attention from investors, developers and technology companies. On September 24, 2026, Nebius (NASDAQ: NBIS) was one of the notable movers, rising roughly 6% intraday as investors responded to developments surrounding AI-compute pricing and new analyst expectations. CoreWeave (NASDAQ: CRWV), another major AI-cloud infrastructure company, was moving in a different direction during the latest trading session, demonstrating just how quickly sentiment can change in this highly volatile sector.
The story becomes even more interesting when leveraged exchange-traded funds are added to the equation.
The Leverage Shares 2X Long NBIS Daily ETF (NEBX) is designed to provide approximately twice the daily percentage movement of Nebius, before fees and expenses. Similarly, the Leverage Shares 2X Long CRWV Daily ETF (CRWG) targets approximately twice the daily percentage movement of CoreWeave. These products are therefore highly sensitive instruments: when the underlying stock moves sharply, the corresponding daily leveraged ETF can move substantially more.
That financial-market activity reflects something larger happening underneath: the rapid construction of the physical and digital infrastructure required to make artificial intelligence work.
What Does Nebius Actually Do?
Nebius is an AI-cloud infrastructure company headquartered in Amsterdam and listed on Nasdaq under the symbol NBIS.
Nebius describes itself as an AI cloud company providing infrastructure across the AI lifecycle, including data, model training, tuning, production runtime and deployment. Its platform is designed around large-scale GPU computing rather than being a conventional general-purpose cloud.
That distinction is important.
Training sophisticated AI models requires enormous amounts of computing power. Once models are trained, they still need computing resources every time users interact with them. This second phase, known as inference, can also require substantial GPU capacity.
Nebius therefore operates in an infrastructure layer underneath many AI applications.
One particularly interesting development is Nebius’s move toward an asset-light infrastructure partnership model. Under this model, infrastructure partners can finance and own data-center hardware while Nebius supplies architecture, software, hardware design and customer demand. Nebius can then market the resulting compute capacity through its sales organization.
That model creates an interesting bridge between AI software and physical infrastructure.
Nebius has also announced major relationships involving NVIDIA and Meta. Its NVIDIA relationship is intended to help scale its AI infrastructure dramatically, while its Meta agreement involves billions of dollars of dedicated AI capacity over multiple years.
What Does CoreWeave Do?
CoreWeave operates in a closely related part of the AI economy.
CoreWeave describes itself as a cloud platform purpose-built for artificial intelligence. Its infrastructure combines GPU computing, networking, storage, orchestration and software specifically optimized for demanding AI workloads.
Instead of an AI company having to purchase enormous quantities of GPUs, construct data centers, build networking infrastructure and operate the entire environment itself, CoreWeave provides access to this infrastructure as a cloud service.
That means an AI developer can effectively rent access to extremely powerful computing infrastructure.
CoreWeave supports workloads including AI model training, fine-tuning and high-throughput inference. Its platform is designed around NVIDIA GPU infrastructure and large-scale clusters.
The company has also been expanding into applications where security and specialized infrastructure matter. In July 2026, CoreWeave announced a collaboration with Leidos involving secure AI cloud capabilities for national-security and intelligence applications.
This is particularly relevant to the broader idea that AI does not necessarily replace professionals. Instead, AI infrastructure can provide the computing foundation that allows professionals to use increasingly sophisticated tools.
Why Are 2X ETFs So Volatile?
The 2X products are a separate part of the story.
A product such as NEBX attempts to deliver approximately twice the daily performance of NBIS, while CRWG attempts to deliver approximately twice the daily performance of CRWV.
For example, if an underlying stock rises 5% in one trading session, a 2X daily product is designed to target roughly a 10% gain for that day before fees and other effects.
The reverse is also true.
A 5% decline in the underlying can correspond to approximately a 10% decline in the leveraged product.
There is another important consideration: daily leveraged ETFs reset their exposure each day. Consequently, their performance over weeks or months can differ substantially from simply taking twice the underlying stock’s cumulative return.
That makes these instruments very different from simply owning NBIS or CRWV.
Why Nebius and CoreWeave Are Interesting Together
Nebius and CoreWeave are competitors in some areas, but they also illustrate a larger AI infrastructure ecosystem.
Think of the AI industry as a stack.
At the bottom are:
Electricity → Data Centers → Cooling → Networking → GPUs → Cloud Platforms → AI Models → Applications → End Users
Nebius and CoreWeave operate primarily in the infrastructure and cloud layers.
Companies building AI applications sit further up the stack.
This creates enormous demand for new data centers, power infrastructure, cooling systems, networking and GPU capacity.
And that brings the discussion directly into an area being explored by Strategic Crypto Reserve (SCR).
Where Strategic Crypto Reserve Fits
Strategic Crypto Reserve has increasingly focused part of its business around AI infrastructure economics and planning software.
The SCR AI Build Configurator is an embeddable planning tool designed to model AI and HPC data-center construction economics. It allows users to examine different configurations, jurisdictions, cooling architectures and accelerator classes while calculating items such as facility load, capital expenditures, energy costs, revenue structures, EBITDA, margins and simple payback.
This creates an interesting conceptual connection with the business models of companies such as Nebius and CoreWeave.
Nebius and CoreWeave need physical infrastructure capable of supporting enormous quantities of AI computing.
The SCR configurator is designed to help people screen and understand the economics of building that infrastructure.
It is not an AI-cloud competitor to Nebius or CoreWeave. Instead, it occupies a completely different layer: infrastructure planning and economic modelling.
A Potentially Interesting Business Ecosystem
This is where the SCR affiliate concept becomes interesting.
The basic idea is that the AI economy is becoming an ecosystem rather than a single product.
An AI company may need GPUs.
A GPU-cloud company needs data centers.
A data center needs electricity.
A data center developer needs financial modelling.
A site-selection consultant needs infrastructure assumptions.
An enterprise needs an AI solution.
Security professionals need tools that can help them analyze information faster while retaining human oversight.
And ordinary businesses increasingly need AI capabilities integrated into their existing operations.
SCR’s stated direction around AI infrastructure therefore sits within a much larger transformation.
AI as a Complement to Professionals
One of the most interesting aspects of this transformation is the possibility that AI will increasingly complement professional work rather than simply replace it.
Security professionals, engineers, analysts, developers, investigators, architects, financial professionals and business operators can potentially use AI to process information, identify patterns, automate repetitive tasks and provide decision-support tools.
The human professional can remain responsible for judgment, context, accountability and final decisions.
The computer supplies scale.
That philosophy fits naturally with the infrastructure being built by companies such as Nebius and CoreWeave.
Without enormous quantities of computing infrastructure, sophisticated AI systems cannot operate at scale.
And without practical tools for deploying that infrastructure, the computing capacity itself cannot easily become useful business capability.
The Bigger Picture
The September 24 market activity surrounding NBIS demonstrates how closely financial markets are watching AI infrastructure demand. Nebius has been discussing higher AI-compute pricing, while the company continues expanding its infrastructure and partnership model.
CoreWeave represents another major version of the same underlying infrastructure thesis: build highly specialized AI cloud infrastructure and sell access to organizations that need enormous computing capacity.
The 2X ETFs add another layer of market exposure, but their daily-reset structure means they should be understood as specialized leveraged trading instruments rather than simple long-term substitutes for the underlying stocks.
For Strategic Crypto Reserve, the more interesting connection may be the infrastructure itself.
The AI revolution requires far more than clever software.
It requires power, buildings, cooling, networking, GPUs, cloud platforms, economic modelling and ultimately applications that improve how people work.
That is the common thread connecting the rapidly expanding AI-cloud infrastructure industry with the SCR AI Build Configurator.
The future of AI will ultimately be measured not only by how intelligent the models become, but by how effectively society can build the infrastructure necessary to put that intelligence to work.
And that is where companies such as Nebius and CoreWeave — alongside infrastructure developers, software providers and planning platforms such as SCR — occupy different but potentially complementary positions in the rapidly developing AI economy.
This article is for informational purposes only and is not financial advice. Leveraged ETFs can be substantially more volatile than their underlying securities, and daily leveraged performance can diverge significantly from longer-term expectations.
How Nebius and CoreWeave would compliment the Strategic Crypto Reserve directives