Home / News / Breaking / Is Green Mining in Trouble? 5 Reasons For Bitfarms Pivot To AI
Breaking 3 min read Add as a preferred source on Google

Is Green Mining in Trouble? 5 Reasons For Bitfarms Pivot To AI

Bitfarms (BITF)

Bitfarms, a company specialised in green BTC mining, has placed its future on a bold new path, announcing a full exit from mining by 2027 to focus on AI and high-performance compute centers. The company plans to repurpose its North American infrastructure, beginning with an 18 MW Washington site retrofitted for GB300-class GPUs. Investors responded immediately. BITF shares fell, reflecting market uncertainty over costs, timing, and a fresh strategy.

The firm’s Q3 results added more context. Revenue reached $69 million, but the company reported a $46 million net loss, partly from winding down Latin American operations. Management paired the earnings release with a detailed plan to convert sites for AI workloads, signaling that Bitfarms sees opportunity in a more predictable, scalable sector.

Why Bitfarms Chose AI Now

The move is not random. Bitfarms faces five economic realities shaping the pivot:

1. Post-Halving Pressures

Bitcoin’s 2024 halving reduced rewards from 6.25 to 3.125 BTC per block. It caused an increase in network activity and a global hashrate above 700 EH/s, which subsequently added mining costs. Even low-cost, renewable-heavy sites struggled to maintain margins.

2. Rising Operating Demands

Bitcoin network difficulty hit record levels, forcing continuous ASIC upgrades. Depreciation and capital intensity mounted, driving quarterly mining costs to $69 million despite revenue growth. Hydroelectric power helped, but hardware refreshes absorbed capital at an unsustainable pace.

3. BTC Price Variability

Bitcoin traded between $96,000 and $115,000 during the months of October and November. Volatile prices magnified revenue swings. Even miners with cheap electricity saw profits fluctuate weekly, showing a challenge that green energy alone cannot solve.

4. Q3 Earnings Miss

Q3 revenue grew 156% year-over-year, yet results still fell short of expectations. Losses from Latin American operations added pressure, and the $46 million net shortfall triggered an 18% drop in BITF shares, reflecting investor frustration with the company’s mining performance.

5. AI’s Stronger Revenue Potential

AI workloads can generate $0.50–$1.00 per kWh versus mining’s $0.02–$0.04. Management predicts the Washington site alone could outperform any previous year of Bitcoin mining. Multi-year GPU contracts offer consistent, predictable returns.

How Bitfarms is Positioned For AI

Bitfarms enters the AI era with 341 MW of active capacity across twelve North American sites and a broader 2.1 GW footprint. That infrastructure allows modular retrofits and liquid-cooled, high-density racks, with Washington leading the charge. A $128 million equipment agreement funds the site, scheduled for completion in December 2026, while Pennsylvania’s Panther Creek campus follows as the next GPU deployment hub.

Financially, the company is well-equipped. Convertible notes totaling $588 million, $300 million in Macquarie-backed project financing, and $230 million in cash and BTC provide roughly $1 billion to accelerate buildouts and future expansions.

Peers like Marathon, Hut 8, and Iris Energy are testing GPU services, but Bitfarms is among the first major public miners to commit to a full, company-wide BTC exit timeline. Its renewable-heavy portfolio, particularly hydroelectric sites, remains a draw for AI tenants seeking low-carbon compute.

What Investors Should Expect

Bitfarms’ shift offers steadier revenue, higher per-kWh margins, and reduced exposure to Bitcoin’s volatility. It stands to generate over 80% of revenue from HPC by 2027, potentially re-rating shares 2–3x if flagship sites secure hyperscaler contracts.

Washington energization, Panther Creek upgrades, and early tenant deals will determine how swiftly the company transforms from green miner to AI infrastructure player, showcasing strategy, adaptation, and the changing economics of power and compute.

Was this Article helpful? Yes No Add as a preferred source on Google
Thank you for your feedback. 0% 0% Add as a preferred source on Google