You are mistaken if you think Soluna Holdings' Q2 revenue growth of 145% tells a story of operational success. The real story is written in the share count: 102.5 million at year-end 2025, 244.6 million by August 10, 2026. A 139% increase in seven months. The ledger remembers what the mempool forgets—dilution is the ultimate tax on the naive.
Context: Soluna is a renewable-powered data center operator that started in Bitcoin mining and is now pivoting to AI infrastructure. The pivot is fashionable. Public miners are selling BTC treasuries, issuing debt, and spinning up GPU clusters. But fashionable doesn't mean profitable. The company reported $15.1 million in Q2 revenue, up 73% excluding a pass-through electricity cost gimmick that added $4.4 million to both revenue and cost of revenue—zero net effect on gross profit. The accounting trick is transparent. The real numbers are not.
Core: The Dilution Engine
Let me walk through the data. I have audited similar pivot stories during the 2021 mining boom. The arithmetic rarely works out for retail when the share count triples. Here, we have a forensic breakdown:
- Revenue growth, but gross profit collapse: Consolidated gross profit fell 60% from Q1 to $766,000. The company blames $1.5 million in maintenance costs at the recently acquired Briscoe Wind Farm, ramp costs at Project Kati 1, and depreciation starting before revenue catches up. That is a polite way of saying the company is burning cash faster than it can generate gross margin.
- GAAP net loss widened to $22.6 million from $17.9 million in Q1 and $7.8 million a year earlier. The loss includes a $4.2 million loss on debt extinguishment. Debt extinguishment is a polite term for “we had to refinance on worse terms.”
- Share issuance is the real business model: In the first half of 2026, Soluna sold 74.2 million shares through its at-the-market (ATM) program, netting $113.5 million. Another 10.2 million shares under a standby equity purchase agreement netted $18.9 million. That is $132.4 million of equity sold in six months. Meanwhile, operating cash burn was $11.6 million, investing outflows were $65.1 million (including $51.4 million for Briscoe), and $25.3 million for interests in Dorothy 1A and 1B. The company is funding its capital expenditures entirely by printing shares.
- Post-quarter dilution continued: As of August 10, Soluna sold an additional 18.8 million shares for $23.6 million, bringing the outstanding count to 244.6 million—139% above year-end 2025. The company is selling shares to cover operating losses, not just growth.
The Pipeline Mirage
Now, the headline: Soluna claims a 6.3 GW data center pipeline. But only 192 MW, or roughly 3%, is operating across three fully energized sites. Another 14 MW is under construction at Kati 1. The remaining 6.1 GW is in planning, development, or assessment. 4.5 GW is in “assessment with power partners.” Assessment is a corporate term for “we have a signed letter of intent, not a shovel-ready project.”
Project Kati 2, a joint venture with Metrobloks, calls for 100 MW of critical IT capacity in phase one and 250 MW in phase two. Neither phase was included in operating capacity. The gap between a press release and a powered rack is vast. I have seen this pattern in the 2021 NFT wash-trading audits—promises of floor price support that evaporated when liquidity dried up.
Floor prices are just liquidated confidence. The same applies to development pipelines. A 6.3 GW pipeline with 3% operational is not a pipeline; it is a wish list. The company is selling shares to fund the assessment phase, which may never convert to revenue.
Contrarian: What the bulls got right
I will give credit where it is due. The company’s revenue growth, even excluding the pass-through adjustment, is real. Project Kati 1 recorded its first positive site gross profit of $82,000. Project Dorothy 1A generated $2.9 million in revenue and $795,000 in gross profit. These are operating assets generating cash flow. The company is building real infrastructure, not just buying GPUs to scalp.
Moreover, the AI demand for data center capacity is real. Wall Street pays up for miners who pivot to AI. VanEck noted that AI-linked miners earn premium valuations before most leased capacity is delivered. Soluna is positioning itself in that narrative. The 6.3 GW pipeline, if even 10% becomes operational, could justify the current valuation.
But the bull case ignores the arithmetic of dilution. A 139% increase in shares in seven months means that even if the company doubles its EBITDA, earnings per share are halved. The market is pricing in an execution miracle that is mathematically improbable. The company will need to raise more capital to build the remaining 97% of its pipeline. The only source of capital is further equity issuance or debt. Both will dilute or burden existing shareholders.
Takeaway: The illusion persists until the liquidity dries
Soluna’s story is written in shares, not hashrate. The company has $15 million in quarterly revenue and a market cap that implies a 30x revenue multiple despite a net loss of $22 million. The only way that multiple holds is if the market believes the pipeline will convert. But the conversion rate from assessment to operating is historically low. I have audited enough balance sheets to know that when the next bear market cycle hits, the companies with the most dilution will be the first to fail.
The ledger remembers what the mempool forgets. The current mempool is full of Soluna hype. The ledger shows a share count that tripled and a net loss that widened. The question is not whether the company can build AI infrastructure. It is whether the existing shareholders will survive the dilution required to build it.
Truth is a derivative of transparent data. The data says: 3% operational, 139% dilution, negative gross margin on a consolidated basis. The narrative says: 6.3 GW pipeline, AI pivot, revenue growth. Choose your derivative.