The AI Power Bottleneck: How Unusual Options Activity Uncovered a Potential Hyperscaler Catalyst

Executive Summary

At SetYourStop, unusual options activity is not automatically treated as a signal to trade the options. We use the data to identify stocks that deserve further investigation.

On Friday, our scanner flagged notable bullish positioning in CleanSpark, Inc. (NASDAQ:CLSK). Approximately 18,839 contracts traded in the $13.50 calls with only seven days remaining until expiration. The calls were out of the money with the stock near $12.85, and net premium reached approximately $462,774. Additional out-of-the-money call buying appeared near the close.

That activity earned CleanSpark a place in the SYS Weekend Report, but the options were only the starting point. Our due diligence uncovered a company that has spent years assembling one of the most strategically relevant assets in the AI infrastructure race: large-scale, grid-connected power.

CleanSpark controls approximately 1.8 gigawatts of contracted power across a portfolio of data-centre sites. Its Sandersville, Georgia campus provides 250 MW of existing energized capacity. Management has said it is in late-stage discussions with high-credit-quality hyperscale tenants, has reached a basis-of-design conclusion with an end-use customer at Sandersville, and is finalizing the structure of a potential lease.

Separate industry reporting has identified Meta as a company discussing capacity at Sandersville. Neither CleanSpark nor Meta has announced an agreement, and the prospective customer has not been publicly identified by management.

The opportunity is meaningful even without assuming who the tenant may be. The confirmed facts show that CleanSpark owns scarce power infrastructure, has an energized site under advanced commercial consideration, and is negotiating with hyperscale customers while demand for AI capacity continues to strain the electrical grid.

A completed Sandersville lease could introduce an entirely different valuation framework for the company. Elevated short interest and concentrated short-dated call positioning could then amplify the market’s response through short covering and dealer hedging.

This is not a prediction that an agreement will be announced. It is a research thesis built from unusual positioning, management commentary, infrastructure scarcity and an identifiable commercial catalyst.

Following the Data

The options activity raised a simple question:

Why would market participants commit nearly half a million dollars to out-of-the-money calls with only seven days remaining?

Short-dated calls leave little time for a thesis to develop. The buyer needs the stock to move quickly enough to overcome both the distance to the strike and rapid time decay. That does not reveal the trader’s motivation, but it can indicate an expectation for near-term volatility.

We then reviewed management interviews, operational commentary, infrastructure assets and developments across the AI data-centre market.

The research moved quickly away from Bitcoin and toward power.

CleanSpark CEO Matthew Schultz has described the company’s accumulated land and power assets as a competitive moat. Management says the company operates approximately 35 sites across the United States, with locations suited to different AI and high-performance-computing requirements.

Metro Atlanta properties can potentially support workloads where latency and fibre connectivity matter. Larger rural sites can support power-intensive training and inference applications.

CleanSpark’s current operations also give it an unusual development advantage. The company can monetize electrical capacity while negotiating with potential tenants and developing purpose-built AI facilities. In Schultz’s description of the strategy, mining funds the infrastructure platform while AI provides the opportunity to monetize it differently.

The market has historically viewed CleanSpark through the valuation framework applied to Bitcoin miners. The company is now positioning its land, utility relationships and power portfolio as infrastructure that could serve some of the world’s largest technology companies.

Why 1.8 Gigawatts Matters

AI infrastructure is no longer constrained only by access to advanced chips.

Hyperscalers also need land, substations, transmission capacity, utility agreements, fibre connectivity, cooling infrastructure and regulatory support. Securing all of those components at the same location can take years.

The third-party analysis reviewed during our research estimated that completing the land acquisition, permitting and high-voltage transmission work required for a major new power site can take approximately five to seven years. The exact timeline varies by jurisdiction, but the larger point is clear: grid-connected capacity cannot be recreated quickly.

CleanSpark says it already has 1.8 GW under contract.

For perspective, the analysis described one gigawatt as enough electricity to serve roughly 750,000 homes under the assumptions it used. That comparison is not a universal conversion, since residential consumption varies by location and methodology, but it illustrates the extraordinary scale involved.

The more important distinction is between planned power and energized power.

A proposed campus may advertise hundreds of future megawatts while still requiring transmission work, utility upgrades, and regulatory approvals. Energized capacity is already connected and operating.

That is why Sandersville matters.

The campus contains approximately 250 MW of existing energized capacity. Management has reported strong customer interest and has progressed through technical planning with an end-use customer.

In an industry where development timelines can determine whether a hyperscaler deploys capacity now or years from now, existing energized infrastructure carries a scarcity premium.

Sandersville and the Hyperscaler Opportunity

During a recent interview, Schultz said CleanSpark was in late-stage discussions with high-credit-quality hyperscale tenants.

He then provided a direct update on Sandersville. Management had received substantial inquiries, reached a basis-of-design conclusion with an end-use customer and was finalizing the proposed lease structure.

A basis of design establishes the major technical requirements for a project. It helps define how the facility must be engineered, configured and developed for the customer’s intended workloads. Reaching that stage indicates a more advanced process than simply marketing available land and power.

The prospective customer has not been publicly named.

Separate infrastructure reporting has connected Meta to discussions involving CleanSpark’s 250 MW Sandersville site. That reporting is relevant because it aligns with management’s confirmation that an end-use customer is already engaged in advanced technical and commercial work. It remains third-party reporting rather than a confirmed agreement.

The strategic fit is understandable.

Hyperscalers are committing enormous capital to AI infrastructure and require large blocks of power that can be deployed faster than traditional greenfield projects. Meta has outlined an aggressive expansion of compute infrastructure, while CleanSpark controls energized capacity in a location management says is supported by the surrounding community and utility relationships.

Schultz has repeatedly emphasized that CleanSpark’s 1.8 GW is spread across jurisdictions that want data centres. He has also said that some communities are offering additional incentives to attract suitable hyperscale tenants.

This local support matters. Data-centre projects can face political resistance, permitting delays and uncertainty over their effect on regional power systems. A site with operating infrastructure, established community relationships and local support can reduce several layers of execution risk.

The current evidence does not establish the identity of the Sandersville customer. It does establish that negotiations have progressed, the power is already energized, and the site is being designed around an end user’s requirements.

The Potential Re-Rating Catalyst

A completed Sandersville agreement would be more than another corporate contract.

It could validate CleanSpark’s transition into AI infrastructure and force the market to reconsider how the company’s power portfolio should be valued.

Bitcoin miners are generally valued using variables tied to cryptocurrency prices, production economics and mining capacity. AI infrastructure companies can attract valuations based on contracted revenue, tenant quality, lease duration, development pipelines and the strategic value of their power assets.

A lease with a high-credit-quality hyperscaler would introduce those factors into the CleanSpark valuation framework.

It would demonstrate that the company’s electrical infrastructure can support institutional-grade AI workloads. It could also create a reference transaction for the remaining power portfolio, allowing investors to assign greater value to other sites capable of future commercialization.

Sandersville is therefore important beyond its 250 MW.

It could establish proof of concept for the broader 1.8 GW platform.

Management has not announced the economics of a potential agreement, so it would be premature to calculate its effect on earnings or cash flow. The immediate significance would be validation. The market would gain a tangible benchmark for assets that have historically been valued primarily through the lens of mining.

That is why a Sandersville deal has the potential to re-rate the stock quickly.

Short Interest, Call Positioning and Squeeze Potential

Market positioning could magnify the reaction to any material announcement.

Recent short-interest reporting placed approximately one-third of CleanSpark’s float in short positions, with close to four days required to cover at average trading volume. High short interest does not create a catalyst, but it can add substantial demand once a catalyst shifts market expectations.

Short sellers who decide to reduce exposure must purchase shares. When a large percentage of the float is already sold short, covering can intensify an upward move.

The options structure introduces a second potential source of demand.

Friday’s unusual call buying established concentrated exposure at the $13.50 strike. Market makers who sell those calls often hedge their risk by buying the underlying stock. If the share price moves closer to the strike, they may need to buy even more shares. Combined with roughly one-third of the float sold short, that added demand could amplify a positive move if a meaningful catalyst emerges.

That process can contribute to a gamma squeeze.

The strongest version of the setup would involve three sources of demand arriving together:

1. Investors buying the stock in response to a meaningful Sandersville announcement.
2. Short sellers purchasing shares to cover bearish positions.
3. Dealers increasing hedges as out-of-the-money calls move closer to or into the money.

The options activity does not prove that an announcement is approaching. It does show that meaningful capital was positioned for a near-term move at a time when management had already disclosed advanced commercial discussions.

Someone always knows something, but an unusual trade does not tell us who knows what. It tells us where to look.

What Could Change the Thesis

The central risk is execution.

Late-stage negotiations do not guarantee a signed lease. The prospective tenant may change its requirements, delay the project or select another location. The cost of converting existing infrastructure into a purpose-built AI data centre could also be substantial.

A Bitcoin mining site cannot be transformed into an AI facility simply by replacing one type of machine with another. AI workloads require advanced networking, fibre connectivity, higher building standards, specialized cooling, and infrastructure designed around expensive GPU clusters.

The identity of the Sandersville customer also remains unconfirmed. Meta may be involved, another hyperscaler may be involved, or the reported discussions may not result in a transaction.

The options activity also has alternative explanations. The calls could represent a directional bet, a hedge against a short position, part of a broader portfolio strategy, or activity unrelated to corporate news.

Those possibilities do not invalidate the research. They define the limits of what the current evidence can establish.

Research Conclusion

CleanSpark entered our Weekend Report because the unusual options activity led to a credible and potentially material research thesis.

The company controls approximately 1.8 GW of contracted power during a period when grid-connected capacity has become one of the scarcest resources in artificial intelligence. Its 250 MW Sandersville campus is already energized. Management says it has completed basis-of-design work with an end-use customer, entered late-stage hyperscaler discussions, and is finalizing potential lease terms.

Industry reporting has connected Meta to the site, although no customer or agreement has been confirmed.

A successful Sandersville commercialization could validate the strategic value of CleanSpark’s infrastructure and introduce a new valuation framework for the wider power portfolio. With roughly one-third of the float sold short and significant short-dated call positioning already present, a meaningful announcement could produce both a fundamental re-rating and a technically amplified response.

It could also lead nowhere.

That is why we conduct research rather than make predictions.

The options activity identified the clue. Due diligence uncovered the infrastructure, management commentary, and potential catalyst. The chart now provides a framework for monitoring the setup as it develops on the right side of a base.

Our objective is to place the opportunity on the radar before an outcome becomes obvious. Readers can then decide whether the thesis fits their own strategy, identify the price levels that matter, and define risk before taking action.

Sometimes these clues precede meaningful developments. Sometimes they amount to nothing.

The process remains the same.

Follow the data. Someone always knows something.

Disclaimer

SetYourStop.com is a research company. We do not tell members what to buy or sell, nor do we make predictions about future market events.

The purpose of this report is to demonstrate how we use unusual options activity, management commentary, industry developments, and technical analysis to identify companies that deserve further investigation. Our goal is to place high-quality ideas on the radar before they become widely recognized, allowing members to build their own watchlists, develop a plan and manage risk according to their individual investment objectives.

The information contained in this report is based on publicly available information believed to be reliable at the time of publication. Discussions regarding potential commercial agreements, market reactions, or valuation changes represent research observations and possible outcomes, not guarantees or predictions. Any reported discussions involving third parties have been attributed to publicly available sources and should not be interpreted as confirmation of a future transaction.

Every investor is responsible for conducting their own due diligence and making their own investment decisions. Markets are inherently uncertain, and every investment carries risk. The responsibility for how this research is used rests entirely with the individual investor.