Company Profile
TeraWulf owns and operates US digital infrastructure, historically tied to bitcoin mining and now repositioning toward power-backed HPC hosting. In Q1 2026 it said it generated $34.0M of revenue, including $21.0M of HPC lease revenue, with 60 MW of operational critical IT HPC capacity at Lake Mariner TeraWulf Q1 2026 results.
Cipher Digital develops industrial-scale data centers for bitcoin mining and HPC hosting. Cipher says its current strategy is to secure power, develop data centers purpose-built for HPC workloads and lease capacity to hyperscale customers; Q1 2026 revenue was $35M and adjusted EBITDA was negative $48M Cipher Q1 2026 business update.
| Company | What it sells | Customer proof | Why it matters |
|---|---|---|---|
| WULF | Critical IT power capacity, data-center shell / hosting infrastructure, and legacy digital-asset mining. | Core42 capacity already generating revenue; Fluidstack / Google-backed Lake Mariner and Abernathy leases TeraWulf Q1 2026 results. | More operating HPC revenue today, but still heavy construction and financing exposure. |
| CIFR | Large-scale data-center campuses, turnkey space and power, and residual mining exposure. | AWS 15-year lease, Fluidstack / Google lease, and a third hyperscale lease announced in Q1 2026 Cipher Q3 2025 business update Cipher Q1 2026 business update. | Cleaner hyperscale data-center landlord story, but current revenue still mostly mining until leases ramp. |
China Competitors / Substitution
China has strong IDC operators and cloud platforms, but it does not provide a full public-market substitute for a US power-backed campus owner signing US hyperscaler AI/HPC leases with Google-, AWS- or Fluidstack-linked counterparties. The closest China comps are segment substitutes: domestic IDC capacity, cloud AI demand, and power/cooling execution.
| Chinese peer / segment | Overlap | Evidence | Substitution boundary |
|---|---|---|---|
| GDS Holdings | High-performance data centers in China and Asia. | GDS reported 2025 net revenue of RMB11.43B and adjusted EBITDA-style data-center economics in its FY 2025 results GDS FY 2025 results. | Large IDC platform, but customer jurisdiction, grid regime, contract law, tenant mix and US hyperscaler access differ from WULF/CIFR. |
| VNET | Wholesale and retail IDC capacity in China. | VNET reported 907 MW wholesale capacity in service and 869 MW committed as of March 31, 2026 VNET Q1 2026 results. | Comparable on capacity delivery, less direct on US hyperscaler AI lease economics. |
| Alibaba Cloud / other Chinese cloud operators | AI cloud demand and domestic compute utilization. | Alibaba Cloud said external cloud revenue growth accelerated to 40% YoY and AI-related products accounted for 30% of external revenue Alibaba Cloud FY2026 update. | Cloud demand validates AI compute growth, yet it is not the same investable exposure as owning US power sites leased to hyperscalers. |
Conclusion: China offers partial substitution in IDC and cloud capacity. For the specific WULF/CIFR thesis, the investable bottleneck is US power, interconnection, construction financing and hyperscale lease execution.
Executive Summary
The analysis view is conditionally positive: WULF and CIFR are credible public-market ways to express the AI data-center power bottleneck, but the equity case depends on execution, financing cost and contract-to-cash conversion. The post's direction is supported by official disclosures; the unsupported part is any shortcut from headline TCV to equity value.
| Item | WULF | CIFR |
|---|---|---|
| Current stance | Positive on operating progress; sensitive to debt, dilution and project delivery. | Positive on hyperscale lease proof; more back-end loaded cash conversion. |
| Current revenue base | Q1 2026 revenue $34.0M; HPC lease revenue $21.0M TeraWulf Q1 2026 results. | Q1 2026 revenue $35M; adjusted EBITDA negative $48M Cipher Q1 2026 business update. |
| Liquidity | Approx. $3.1B cash and restricted cash at Q1 2026 quarter-end TeraWulf Q1 2026 results. | Company says it has sufficient capital for the third campus equity component and near-term needs; Q1 release also notes a $200M revolver Cipher Q1 2026 business update. |
| Next proof point | CB-3/CB-4/CB-5 and Abernathy delivery through 2026. | Barber Lake, Black Pearl and third-campus lease milestones through 2026-2027. |
Key Takeaways
- WULF has moved beyond a pure bitcoin-mining narrative: Q1 2026 HPC lease revenue already contributed $21.0M TeraWulf Q1 2026 results.
- CIFR has stronger named hyperscaler validation from AWS plus Fluidstack/Google, but cash flow still lags lease announcements until delivery ramps Cipher Q3 2025 business update Cipher FY 2025 business update.
- Power is the scarce input. Uptime Institute's 2025 survey highlights rising costs, power constraints and AI demand challenges across the data-center industry Uptime Institute 2025 survey.
- Headline contract value should be converted into ownership-adjusted revenue, lease start dates, financing costs, maintenance capex and share dilution before drawing an equity conclusion.
- China has strong domestic substitutes for IDC and cloud capacity, but the US hyperscaler lease / US grid bottleneck is not fully replicated by Chinese listed peers.
Recent Two-Year Share Price Trend
Yahoo Finance chart API; normalized monthly last closes. WULF: 2024-07-03 $5.21 to 2026-07-02 $21.18, +306.5%. CIFR: 2024-07-03 $4.65 to 2026-07-02 $20.04, +331.0%. Yahoo Finance WULF chart API Yahoo Finance CIFR chart API
The chart matters because both stocks have already rerated with the AI infrastructure narrative. New capital is no longer buying a hidden mining-to-AI pivot; it is underwriting delivery, financing cost and tenant-credit proof after a large move.
Market Context
AI training and inference workloads push demand toward dense power, cooling, interconnection and site execution. In this market, the bottleneck is not only GPUs; it is the ability to turn land, grid access, substations, financing and construction into usable critical IT load. Uptime Institute's 2025 survey describes the sector as facing rising costs, worsening power constraints and AI-related challenges Uptime Institute 2025 survey.
| Driver | Positive read-through | Counter-evidence to monitor |
|---|---|---|
| Hyperscaler demand | AWS, Google-linked Fluidstack and other customers are signing long-duration capacity agreements. | If hyperscalers self-build faster or renegotiate terms, scarcity pricing can fade. |
| Power and interconnection | Sites with ready or near-ready power can command strategic value. | Grid queues, local permitting, equipment lead times and community opposition can push revenue out. |
| Financing | Credit-enhanced leases can support project-level debt. | High coupons, convertibles and equity issuance can transfer value away from common shareholders. |
Company / Segment Quality
| Segment | WULF quality read | CIFR quality read |
|---|---|---|
| HPC leases | Q1 lease revenue shows the model has begun to operate, with 60 MW energized at Lake Mariner TeraWulf Q1 2026 results. | Hyperscale lease proof is strong, but revenue contribution is largely future-dated Cipher Q1 2026 business update. |
| Mining legacy | Provides operational history and site infrastructure but creates commodity exposure and accounting noise. | Still reported bitcoin-mining revenue in Q1 2026, while the brand and capital allocation moved toward data centers. |
| Development pipeline | Targeting 250-500 MW of new contracted capacity annually and adding Kentucky, New York and Maryland power sites TeraWulf Q1 2026 results. | Barber Lake, Black Pearl and Colchis give scale, with Colchis positioned as a 1 GW West Texas site Cipher Q3 2025 business update. |
Orders / Customers / Partnerships
| Proof point | Official status | Investment implication |
|---|---|---|
| WULF - Lake Mariner | Core42 60 MW and Fluidstack / Google-backed capacity at Lake Mariner are disclosed in TeraWulf's FY 2025 and Q1 2026 updates TeraWulf FY 2025 results TeraWulf Q1 2026 results. | Recurring lease revenue is already visible, reducing pure optionality risk. |
| WULF - Abernathy | JV with Fluidstack and Google designed for 168 critical IT MW in later FY 2025 disclosure, with construction targeted for 2026 delivery TeraWulf FY 2025 results TeraWulf Q1 2026 results. | Creates geographic diversification, but joint-venture economics and ownership-adjusted cash flows matter. |
| CIFR - AWS | Approx. $5.5B, 15-year AWS lease to deliver 300 MW in 2026 Cipher Q3 2025 business update. | Best single proof point for Cipher's hyperscale credibility. |
| CIFR - Fluidstack / Google | FY 2025 update says Cipher had 600 MW gross contracted HPC capacity across AWS and Fluidstack / Google leases Cipher FY 2025 business update. | Shows the model is not dependent on one tenant, but construction and debt remain key. |
Competitors
| Peer | Overlap | Relative strength | Key difference |
|---|---|---|---|
| CoreWeave / Nebius | AI cloud and GPU capacity contracts. | More direct GPU-cloud revenue exposure. | Valuations already reflect faster AI-cloud growth; covered separately on learnsomefinance. |
| IREN / Core Scientific | Power-backed miners pivoting to AI/HPC. | Potentially similar power-site optionality. | Different customer proof and capital structures. |
| GDS / VNET | Data-center capacity and Chinese cloud/IDC demand. | Established China IDC scale. | Less direct exposure to US hyperscaler AI lease economics. |
| Traditional data-center REITs | Data-center ownership and leasing. | Lower business-model risk. | Less mining-to-AI optionality and often lower growth torque. |
Financial Statements
| Metric | WULF | CIFR | Read-through |
|---|---|---|---|
| Revenue | Q1 2026 revenue $34.0M, including $21.0M HPC lease revenue TeraWulf Q1 2026 results. | Q1 2026 revenue $35M, still reported as bitcoin-mining revenue in detailed statements Cipher Q1 2026 business update. | WULF currently shows more HPC revenue conversion; CIFR has more future lease conversion. |
| Profitability | Q1 2026 net loss attributable to TeraWulf was $427.6M, affected by warrants, interest, impairment and other items TeraWulf Q1 2026 results. | Q1 2026 net loss was $114.3M and adjusted EBITDA was negative $48.2M Cipher Q1 2026 business update. | Both are investment-stage infrastructure equities, not mature REIT cash-yield vehicles. |
| Cash / funding | Cash and restricted cash were about $3.1B at March 31, 2026 TeraWulf Q1 2026 results. | CIFR cited a $200M revolver and sufficient near-term capital for its third campus equity component Cipher Q1 2026 business update. | Liquidity is substantial, but debt service and construction spend remain central. |
| Financing quality | TeraWulf completed more than $5B of financings in Q3 2025 and $6.5B of committed capital in FY 2025 disclosures TeraWulf Q3 2025 results TeraWulf FY 2025 results. | Cipher completed three bond offerings for aggregate proceeds of $3.73B in FY 2025 update Cipher FY 2025 business update. | Project finance validates the contracts but can cap common-equity upside if returns are debt-heavy. |
Forecast / Operating Bridge
The bridge below is illustrative and should not be read as company guidance. The logic is to track contracted critical IT MW into delivered MW, delivered MW into rent start, rent start into EBITDA, then subtract interest, maintenance capex and dilution.
| Bridge step | Base case marker | Upside marker | Failure marker |
|---|---|---|---|
| Capacity delivery | WULF and CIFR deliver scheduled 2026 phases with small slippage. | Additional 250-500 MW annual contracting for WULF and additional hyperscale leases for CIFR. | Grid or construction delays push rent starts beyond expected windows. |
| Revenue conversion | HPC leases become the majority revenue driver by 2027. | Tenant expansions and annual escalators improve recurring revenue visibility. | Mining remains a large revenue share for longer than expected. |
| Financing | Debt service covered by tenant-backed contracted revenue. | Lower refinancing cost after projects are operating. | High-yield coupons and convertibles absorb most economics. |
Valuation / Scenarios
For WULF and CIFR, EV/revenue based on today's mining revenue is not very informative. A cleaner lens is capacity-based: value delivered and contracted critical IT MW after deducting construction cost, debt, partner economics and dilution.
| Scenario | What must happen | Equity implication |
|---|---|---|
| Downside | Construction delays, higher capex, tenant onboarding slippage, or weak lease economics. | Headline TCV fails to protect the common stock; multiples compress. |
| Base | 2026 projects are delivered broadly on schedule; cash burn is funded without major new equity shock. | Stocks can hold a premium as scarce-power infrastructure platforms. |
| Upside | New customers sign at attractive rents, delivered MW grows faster than debt, and operating EBITDA scales. | Re-rating can continue, with WULF helped by current HPC revenue and CIFR by hyperscale tenant proof. |
Catalysts
- Quarterly disclosure of delivered critical IT MW, rent commencement and HPC lease revenue.
- WULF CB-3/CB-4/CB-5 and Abernathy construction milestones through 2026.
- CIFR Barber Lake and Black Pearl delivery updates, AWS phase delivery and any additional hyperscale lease.
- Debt refinancing, revolver use, convertible issuance, warrant exercise and share-count changes.
- Utility interconnection, power purchase, permitting and local project updates at new sites.
Risks
- Construction risk: delayed substations, transformers, cooling, racks or tenant hardware can defer rent.
- Customer risk: long contracts still include delivery obligations, service levels and counterparty-specific economics.
- Financing risk: high-yield notes, convertibles, warrants and partner JVs can dilute common-equity returns.
- Technology risk: GPU depreciation, model architecture changes or lower cloud prices can pressure tenant economics.
- Commodity risk: residual bitcoin-mining operations keep some exposure to BTC price, hashprice and power markets.
- Regulatory and community risk: large data-center power draws face local, utility and environmental scrutiny.
Questions for Management
- What percentage of contracted MW is already rent-generating, under construction, or still dependent on permitting/interconnection?
- How should investors translate gross contract value into ownership-adjusted revenue and free cash flow?
- What are the minimum take-or-pay, termination, delay and service-level terms by major tenant?
- How much additional equity, convertible debt or project-level debt is required after the current funded plan?
- What maintenance capex and refresh obligations remain with WULF/CIFR versus the tenant?
- How much residual bitcoin-mining exposure will remain once HPC leases ramp?
Social Thesis
Original X/Twitter thesis. The post frames neocloud earnings as evidence that Michael Burry's GPU-depreciation concern is missing the contract and power side of AI infrastructure. The directly actionable under-covered target is TeraWulf / WULF; Cipher Digital / CIFR is a close peer and positive reference point because the post compares WULF's update with Cipher's AWS-backed HPC lease.
Source-quality note: the original X URL is the social source. fxtwitter/vxtwitter were used only to preserve visible text when X public HTML was inconsistent; all material business claims below are checked against company releases and filings-style financial releases fxtwitter social-source mirror vxtwitter social-source mirror.