TeraWulf is a front runner in the Bitcoin mining space following significant news announcements in August that Google is taking an 8% stake in the company through a partnership with European-based neocloud company Fluidstack.
The announcement in mid-August was initially for 200 MW over a 10-year period, for $3.7 billion in contracted revenues with a potential to reach $8.7 billion through lease extensions. Of this, Google backstopped $1.8 billion by taking an 8% stake in TeraWulf.
Days later, it was announced that Fluidstack was increasing its lease option for $6.7B in contracted revenue with a potential of up to $16B in lease obligations. Google also stepped to backstop $3.2 billion for an equity stake worth 14%.
As you can imagine, the stock surged off the initial deal. The strong price action was supported by the second announcement to where the 130% surge in August has held up.
The crux of the issue for Bitcoin miners is not real estate and power, rather the quality and longevity of these stocks is dependent on diversifying beyond cash-strapped neoclouds to improve their credit worthiness. To some extent, neoclouds signing lease deals with Bitcoin miners is a house of cards as both are burning cash. For creditors, this means cash flows could fall short of covering the construction costs by both the lessee building the site and the lessor funding the site.
Therefore, when a Miner secures a hyperscaler, the prospects of raising capital and the terms for raising the capital improve substantially. Miners need access to cheap capital, and having a hyperscaler back their lease deals help to achieve this as it provides committed cash for a lender.
Bitcoin miners that can offer investors visibility in terms of funding the construction of these massive (and rapid) multi-billion dollar projects will inevitably do better than stocks where capital is uncertain (or is coming from a company that also has to raise capital).
Below, we explore how Google’s investment in TeraWulf lays the foundation for more durable growth and paves the way for expanded capacity that could (potentially) drive the stock higher. TeraWulf is the second report in a Series of Bitcoin Miners that we have planned for our premium members over the coming weeks. We’ve also covered a thematic overview here.
Note, this is a momentum stock and we plan to adhere to risk management.
TeraWulf has Roughly $700M in Annual Lease Agreements
In the earnings call, management provided more color on the Fluidstack/Google deal, stating the original $3.7B in contracted revenue would provide $350 million a year in lease revenue with net operating margins at 85%. There was an additional 30-day exclusivity deal on data center site CB-5, which is what led to the second announcement a few days later.
The second announcement essentially doubled the original agreement with $6.7B in contracted revenue. Therefore, one can reasonably expect that TeraWulf will see $700M in annual lease agreements from these two announcements, if we assume similar lease terms as the original announcement.
There are still a few unknowns, such as when the site will be delivered in order to recognize the revenue on the income statement, if the lease extensions of up to $16 billion will be exercised, and how TeraWulf will fund the capex required for the data center sites.
Here are the terms of the deal:
- 10-year for 200-plus MWs were the original deal terms with management stating it was “200 megawatts of critical IT load, about 250 megawatts of gross site capacity.”
- The additional 30-day extension, which was quickly exercised, added another 1602 MW of critical IT load.
- According to management commentary on the initial deal, the deployment will begin in H1 2026 with 40MW and be fully deployed by year-end.
As stated, Bitcoin miners are seeing rapidly improving financials as the deal described above results in an 85% operating margin. In 2022, TeraWulf’s operating margin was awful at (270%) and at (43%) last year. Therefore, the fundamental profile of these deals are providing a sharp rebound that we hope to participate in.
TeraWulf has over 1GW of Capacity Total
TeraWulf has two locations: Lake Mariner with up to 750 MW and a new site Cayuga which offers up to 400MW. According to a previous earnings call, the company has plans to add 250 MW more. Prior to the Fluidstack/Google deal, the primary customer was Core 42, which is an AI data center company headquartered in Abu Dhabi.
The contract discussed above is between Fluidstack and Google at the Lake Mariner site, which is located near Buffalo, New York. There is a new site located in Lansing, New York that offers up to 400 MW on a fully equipped site, “with high-capacity transmission, industrial water intake and redundant fiber.” The Cayuga site will offer 130MW initially by 2027. The amount paid for Cayuga was $95 million in stock with $3 million in cash.
TeraWulf discusses their data centers as acronyms:
- CB-1 will begin to generate revenue this quarter ending in October
- CB-2 will begin to generate revenue in the December quarter
- CB-3, CB-4 and CB-5 will offer a combined 366 MWs and are stated to be on a tight timeline. Later in the call, management discussed their goal is to typically deliver within a year: “So as you think about it, once we sign a deal, we're typically delivering the facilities within a year. So we're already thinking ahead to meet that time line.”
How TeraWulf Separates Itself from other Bitcoin Miners
According to Terawulf, there are a few key reasons Fluidstack and Google chose them over other miners. It’s generally understood that hyperscalers go through intensive site reviews, and thus, there was a question on the call as to what helps Terawulf stand apart.
- The first reason is the attractiveness of the site, which management stated ticks all of the boxes: “our installed energy infrastructure, redundant grid connections, the land, the water, fiber latency, 89% 0 carbon power source, availability power.”
- Secondly, the company has been working with Core42 on liquid-cooled data centers, and can help Google save time by having already navigated this complexity.
- Third, the team has 15 years of experience working on complex infrastructure projects, and is able to confidently execute on these multi-billion dollar projects: “We've got decades of experience executing complex energy infrastructure projects. And I don't think that's something you can replicate overnight.” Building on the “decades of experience executing complex energy infrastructure projects,” TeraWulf announced an acquisition of Beowulf, an energy-powered digital infrastructure company. The deal is worth $52.4 million with $3 million in cash and 5 million in shares. As part of the deal, 94 Beowulf employees will be integrated into the workforce and this will terminate large payments TeraWulf was paying to Beowulf.
A new fourth way that Terawulf separates itself may be the most distinguishing factor, which is that TeraWulf will have better credit terms than other Bitcoin miners as Google is their largest shareholder and is back-stopping the lease by guaranteeing 50% of the lease value. On the call, this piece was stated to be a game changer as access to capital and creditworthiness is arguably the most important piece to help separate Bitcoin miners. Here is what was stated: “You look at Google as a financial partner here working with us, a multitrillion-dollar market cap, one of the largest companies in the U.S. and with a measly $30 billion, I think, of debt or so […] And so I think we will look to finance the site in totality under this new credit regime and new support regime. So I think that's the obvious and clear takeaway.”
Management also emphasized that having access to land and power is the easy part of the execution, whereas financing is the harder piece: “Having a signed deal is more than simply having land and access to power. You have to finance the transaction. It's critical to the magnitude of spend. And that's why it was so important to have Google involved pretty much from the start.”
Investors obviously do not have the ability to do extensive site reviews to distinguish Bitcoin miners. Therefore, hyperscalers will help signal to the market which ones are most attractive, and will help signal to the market which ones will likely be able to execute on any lease deals. Google back-stopping TeraWulf through a deal with Fluidstack is a vote of confidence that we are paying close attention to.
Potential for More Deals:
On the most recent earnings call, management discussed the possibility of expanding to add more sites:
“Finally, regarding our growth pipeline. We are constantly evaluating additional sites to add to the TeraWulf portfolio and maintain an extremely rigorous approach to this process. In 2025, we have evaluated over 75 potential expansion sites, and from that, we have a handful of progressing through negotiations.”
Perhaps the best statement in the earnings call was the goal of TeraWulf to repeat the Fluidstack/Google deal as much as possible, with a high-level overview that 2026 could be better than 2025 in that regard:
“On pricing, we're very happy with the economics from Core42 and the Fluidstack Google deal, and think shareholders will be rewarded if we just keep replicating them. I think there's a good argument that the market might even be tighter in 2026 than in 2025 given ongoing power constraints and rising hyperscaler CapEx.”
Revenue
Revenue trends tell the story of a miner on the verge of a V-shaped rebound. In Q2’25, revenue came in at $47.6 million, up 34% year-over-year and 39% sequentially, as bitcoin prices rose and hashrate climbed to 12.8 EH/s.
The prior quarter was softer at $34.4 million, hurt by Polar Vortex and power cost spikes. Importantly, Q2 marks the last quarter of “BTC-only’ revenue. Beginning in Q3’25, WULF will start recognizing contracted HPC lease revenue from Core42’s 72.5 MW deal.
Management confirmed that “WULF Den” was energized in July, CB-1 followed in August, and CB-2 is on track for Q4. This timing makes Q3 the Company’s first blended quarter of mining and HPC revenues, which should begin to shift investor perception away from speculative BTC exposure toward contracted, more predictable AI infrastructure income.
As seen above, analyst top-line expectations are signaling confidence in the Company’s transition from pure-play bitcoin mining into contracted HPC infrastructure. Forward revenue growth is projected to accelerate sharply into 2026, with consensus now calling for revenue to nearly double YoY in Q3’25 (103.0% YoY), sustaining that elevated growth level through Q2’26 (87.24% YoY). These estimates imply a steep step-up in contribution from the Core42 HPC lease that began in Q3’25, alongside continued hashrate growth on the mining side.
This quarter should be viewed as a financial bridge. Historical results remain tethered to bitcoin economics while forward consensus increasingly reflects the durability of contracted HPC leases. The mix shift from volatile BTC mining to long-dated AI hosting contracts is the key re-rating catalyst here.
Margins
Gross margin improved significantly in Q2 as power costs normalized. Cost of revenue fell from 71.4% of revenue in Q1 to 46.4% in Q2, bringing gross margin back to a healthier 54%. On a year-over-year basis, gross margin was down from Q2’24, reflecting the impact of the April 2024 halving and rising network difficulty. While volatile, sequential improvement in gross margins should be viewed as a clear positive. WULF’s gross margin of 54% in Q2 now looks closer to peer average (IREN / RIOT mid-50’s), showing that the underlying economics are competitive.
Operating performance remains pressured by on-going build costs. While gross margin recovered, operating losses continue, reflecting elevated SG&A and depreciation tied to the Lake Mariner expansion. Operating leverage should improve once HPC lease revenues flows, but in Q2, WULF still posted an operating loss. Operating margin headwinds are largely transitional and tied to scaling corporate overhead in anticipation of HPC ramp
At the net income level, losses narrowed to ($18.4) million in Q2 from ($64.1) million in Q1, still worse than the ($10.9) million in Q2’24. The YoY comparison underscores the ongoing headwinds from the halving and the timing of new infrastructure coming online.
Earnings
EPS dynamics remain noisy due to financing and non-cash accounting charges. On a GAAP basis, EPS losses have been volatile: Q2 GAAP net loss equated to ($0.05) per share, improving from ($0.18) in Q1 but worse than ($0.03) in the year-ago period.
This mirrors the broader group where GAAP EPS is distorted by financing and non-cash charges: APLD shows the same dynamic while IREN is the rare exception already printing GAAP profits due to its GPU cloud mix. For WULF, the cleaner signal is adjusted EBITDA, which turned positive this quarter.
The underlying picture is more stable on an adjusted basis. Adjusted EBTIDA swung positive in Q2 at $14.5 million compared to ($4.7) showing a meaningful improvement in operating performance when one-time and non-cash charges are stripped out. This mirrors the broader pattern across the miner group, where adjusted results are the cleaner signal of progress while GAAP optics remain distorted by convertible debt, accounting, depreciation, and stock compensation.
Analyst estimates paint a very constructive picture for WULF’s EPS trajectory – moving from consistent GAAP losses through FY25 into profitability beginning in FY26. From there, the ramp is steep: consensus sees EPS flipping positive in Q4’26 and growing more than 200% YoY by Q4’27. This progression mirrors the broader miner-to-AI pivot: short term pain, medium term inflection, and long-term structural profitability. Analysts are effectively modeling WULF as a credible AI infrastructure provider by 2026, with the EPS trajectory confirming that transition.
Cash Flow and Balance Sheet
Operating cash flow swung from an inflated $56.5 million gain in Q1’25 to a $54.8 million outflow in Q2. The sharp sequential decline was largely mechanical. Q1 benefited from a one-time $90 million deferred rent prepayment while Q2 represents a more normalized bun rate tied to the company’s build cycle. On a year-over-year basis, the trend also deteriorated, with Q2’25 OCF falling well below the $16.4 million generated in Q2’24, underscoring the heavier working capital load associated with the HPC construction.
Free cash flow outflows widened to $174.8 million in Q2, compared to $37.2 million in Q1 and $30.2 million in Q2 of the prior year. The step down was driven by accelerated capex into Lake Mariner data halls, $120 million in Q2 versus $94 million in Q1, on top of weaker operating inflows. Put differently, WULF’s free cash burn in Q2 was nearly six times higher than the same period last year, highlighting the front-loaded nature of its HPC investment cycle.
Liquidity contracted alongside these outflows. Cash balances fell to $90.0 million at quarter-end, down from $219.6 million in Q1 and $274.5 million in Q4’24, representing a sequential draw of $130 million and a modest $15 million decline year-over-year. Debt held steady at roughly $489 million, leaving the balance sheet pressure solely on cash.
Working capital shifts add nuance to the liquidity story. Accounts receivable rose modestly to $1.2 million from $0.5 million in Q1. With DSO still under 10 days, collections remain swift, leaving receivables immaterial in the broader picture. Accounts payable held flat at $38.8 million sequentially but remained elevated from $24.4 million in Q4’24, suggesting WULF is leaning slightly on vendor float as a temporary source of liquidity while funding its HPC build.
Project Costs and Financing
Cash / Debt Position: At the end of Q2’25, cash stood at $90.0 million and debt at $488.72 million, yielding a Cash / Debt ratio of .18x. This lags peers such as IREN (0.59x) and RIOT (0.38x), while being broadly in line with APLD (0.18x). While liquidity has compressed to $90 million against $489 million of debt, the balance sheet is not deteriorating structurally. Debt has remained flat, and the drawdown is tied directly to front-loaded HPC spend. With Core42 lease revenue beginning in Q3, WULF is positioned to pivot from cash burn to recurring contracted inflows, a trajectory that analysts have started to model into FY26 and beyond.
Conclusion:
With Google now a strategic backer through its Fluidstack partnership, TeraWulf is positioned to access cheaper capital and expand its Lake Mariner facility toward its 750 MW potential. The Cayuga site pushes the company past the 1GW capacity mark, and counting.
While execution risks remain around rapid build-outs and financing, the company stands out as one of the few Bitcoin miners offering investors an opportunity to participate in high-margin AI data center deals.
We hold a small allocation in TeraWulf and plan to actively risk manage the position.
Please note: The I/O Fund conducts research and draws conclusions for the company’s portfolio. We then share that information with our readers and offer real-time trade notifications. This is not a guarantee of a stock’s performance and it is not financial advice. Please consult your personal financial advisor before buying any stock in the companies mentioned in this analysis. Beth Kindig and the I/O Fund own shares in WULF at the time of writing and may own stocks pictured in the charts.
Recommended Reading: