North American AI HPC/DC:
Questions we ask ourselves?
🔹How are retail and wall street analysts pricing GW/price/ticker?
- How do you price near-term GW activation?
- BTC mining aside, are you assigning a multiple to the AI HPC/DC assets to determine over/under valuation?
- If the assets are coming online, how is wall street making risk-adjusted calculations on value?
🔹Is big money finally waking up?
- JPMorgan Launches $1.5 Trillion Plan to Support Industries Deemed Critical to U.S. Interests
- Blackstone to invest $25 billion in Pennsylvania data centers and natural gas plants
- OpenAI, Broadcom Forge Multibillion-Dollar Chip-Development Deal
- Mag10 deals across the board, 10-12 AI HPC/DC deals 2025 YTD provide insight into a framework for contract closures and MW/GW activation.
🔹Key Pockets of Interest?
$IREN &
$CIFR -- two key leaders in the AI HPC/DC arena for crypto/hybrid pivots. The near-term activations is something the street realizes, and they've come around to the terms of "IRENing". The infra efficiency KPIs clearly favor
$IREN -- giving them major flexibility in near and long-term optionality. If you pick one -- you pick
$IREN, if you pick 2, you add
$CIFR. Again -- maximizing profits, some prefer 1 investment, i prefer multiple based on near-term MW/GW activation and over/under valuation.
$BITF -- what was clearly an undervalued play -- late arrivers failed to take into account the opportunity of the high value of locality, infra requirements, and quality. The key in the acceleration of the stock price -- i believe is tied to the quality of energy, criteria for infra, and locality. If you only focused on MW/GW sizing -- i think that was the gap. Regional/regulatory dynamics are considered favorable to
$BITF as well.
$HUT -- What stands out is the near-term YTD decision to go fully all-in by breaking out ABTC and going hybrid crypto/AI-hpc-dc -> this differentiates itself from
$CLSK which is doing a CSP pilot and AI HPC/DC transition, along with their extended BTC business. Though execution history could potentially favor
$CLSK and near-term MW/GW activation favors
$CLSK. The street is now looking forward, where the undervaluation risk-adjusted opportunity -- the further you look out-- is starting to turn in
$HUT's favor.
$HUT also has a hybrid model which is something to watch for moving forward. Arguable
$CLSK has also proven better PUE metrics, but debatably these two are H2H favorites.
$SLNH - Other tickers appear to have hit what wall street sees as their valuation targets -- which leaves only a few smaller players behind. One that is intriguing is
$SLNH -- an undervalued hybrid pivot, but the risk-adjusted valuation keeps the share price low -- given their financial risk and financial instability -- along with their requirement to prove and execute out their business model -- this is one to watch for.
------------------------------
Please note that the following information is an assumptive/test scenario/forecast of what the competitive landscape could look like, and your feedback is helpful in the comments to provide color to the analysis or to point out any mistakes for the broader audience.
NFA - Not Financial Advice
DYOR - Do Your Own Research
AI Usage - Leveraged for data, sourcing, & summarization
------------------------------
## 1. Overview
North American AI HPC/DC sector dynamics in mid-October 2025 highlight accelerating hyperscaler commitments like OpenAI-Broadcom's 10 GW accelerator push and Bloom Energy's $5 billion fuel cell deployments, offset by persistent grid bottlenecks projecting 38 GW new loads by 2030 and vacancy at 1.6%, creating balanced growth prospects amid execution challenges.
------------------------------
## 2. Optimized Framework for AI HPC/DC Analysis
Enhanced for October 2025 patterns like fuel cell adoptions in 25% partnerships and liquid cooling in 73% new AI sites, framework maintains power security at 35% while adding profitability signals (15% weight) via PUE (target <1.2 for efficiency), revenue-cost per MW differentials (target >$10M/MW/year margins), bleeding into valuations for undervaluation flags; closures forecast at 5.7 months with 18% deferral risks from energy costs.
🔹 Power Security (35% weight): Stresses PPAs/on-site solutions like Bloom fuel cells (e.g., Brookfield $5B deal), cutting 20% delays; per Reuters (10/13/2025), Bloomberg (10/13/2025), countered by McKinsey 12% U.S. overuse by 2030.
🔹 Funding Structure (20% weight): Emphasizes debt/equity for 50% growth (e.g., OpenAI-Broadcom 10 GW), per WSJ (10/13/2025), SEC; balanced by Gartner $40B depreciation on 30% efficiency gains.
🔹 Execution and Time-to-Value (20% weight): Targets <12-month ramps, crypto pivots saving 40% (e.g., WULF expansions); per Statista (10/2025), JLL (09/2025), flagged for 10-20% IRR reductions from fees.
🔹 Technology Lock-In and Sustainability (10% weight): Includes GPU/cooling for 100 kW/rack (e.g., ABB-NVIDIA), renewables at 536 TWh; Deloitte (2025) notes 50-60% QoQ efficiency, offset by CBRE 1.9% vacancy.
🔹 Regulatory Clearance and Leadership Strength (10% weight): Factors incentives (e.g., Louisiana Meta) and pivots (e.g., GLXY Helios); Utility Dive (10/2025) on Texas mandates, countered by PJM 30 GW adds.
🔹 Profitability Signals (15% weight): Integrates PUE (AI avg 1.5-1.7, target 1.2), revenue per MW ($12.5M/year AI vs $4.2M traditional), cost per MW ($10-20M build, $0.10/kWh opex); high margins (>80% gross) boost valuations, per Deloitte (2025), CBRE H1 2025; low PUE/revenue-cost gaps flag undervaluation, balanced by 20% power inflation risks.
------------------------------
## 3. Key Entities in North American AI HPC/DC Landscape
Spanning 50 entities with ≥99% share per Statista/Gartner (10/2025), categorized by hyperscalers (70%), colocation (20%), niche (10%), crypto pivots (21% overlap), AI hardware; viability updated for partnerships like OpenAI-Broadcom, with MW 50 to 10 GW, 2025-2026 focus incorporating PUE/cost signals for profitability.
------------------------------
## 4. YTD 2025 Deals and Activations Summary
YTD to October 13, 2025, 14 deals closed at 15.528 GW and $8.61 billion annualized value, with activations like Bloom-Brookfield $5 billion fuel cells stressing sustainability in 65% Q4 announcements, yet 18% deferrals from grids amid 3:1 demand gap; profitability signals show AI revenue $12.5M/MW/year vs $10-20M build costs.
🔹 Key closures: CoreWeave-Meta (300 MW, Q3); Microsoft-Nebius (300 MW, Q3); OpenAI-Oracle-SoftBank (7 GW, Q4); Meta-Entergy (2 GW, 2026); TeraWulf-Fluidstack (360 MW, Q4); OpenAI-Broadcom (10 GW by 2029, Reuters 10/13/2025); Bloom-Brookfield ($5B, DCD 10/13/2025).
🔹 Growth metrics: Volume 600% YoY, 584.3% MW CAGR (code validated), $500M savings per pivot but 20% inflation; McKinsey (10/2025) to 123 GW by 2035, Gartner $1M/min downtime flag.
🔹 Activations online: 6.2 GW YTD (e.g., Centersquare 10 DCs, Baxtel 10/04/2025), 9 GW pipeline 75% preleased per JLL (10/2025); CBRE (09/2025) 43.4% growth but 1.9% vacancy counter.
🔹 Impacts: 12% pipeline boost but 38 GW strains, fuel cells/liquid cooling (73% new sites) cut 50-60% energy QoQ per Deloitte (2025); Statista (10/2025) $400B spend, offset by depreciation.
------------------------------
## 5. Stack Ranking for Near-Term Uncontracted GW/MW Activation
Top 12 ranked by viability ( 0.2 for tech integrations), Q4 2025-H2 2026 uncontracted focus for CSP/colocation/hybrid revenue; CIFR/IREN lead at 95/100 on deals, 2-5x upside vs 15% delays; GW/valuation via EV per GW (peer avg $9.5B/GW) now integrates profitability (PUE 1.5-1.7 avg, revenue $1.78-12.5M/MW/year minus $10-20M costs for >80% margins), flagging undervaluation if high margins/low PUE not priced.
🔹 CIFR (95/100, Crypto Pivot/Hybrid): 300-507 MW Barber Lake TX, Q4 2025; $3B Fluidstack; GW/Valuation Opportunity: EV $5.56B on 0.4 GW implies $13.9B/GW, overvalued but profitability (PUE ~1.5, $1.78M/MW/year revenue-cost >80% margins) justifies 20% premium, undervalued on execution.
🔹 IREN (94/100, Crypto Pivot/CSP): 1.4 GW Sweetwater TX, April 2026; NVDA speculation; GW/Valuation Opportunity: EV $13.19B on 1.4 GW implies $9.4B/GW, fair with 15% upside; profitability (renewables aid PUE <1.4, $12.5M/MW/year potential) signals undervaluation vs peers.
🔹 HUT (95/100, Crypto Pivot/Colocation): 1-1.5 GW TX/LA, Q4 2025 300 MW; savings; GW/Valuation Opportunity: EV $4.27B on 1.25 GW implies $3.4B/GW, deeply undervalued (64% below); profitability (PUE 1.5, $0.5-0.8M/MW/year HPC) boosts 2-3x re-rating potential.
🔹 WULF (90/100, Crypto Pivot/Hybrid): 300-400 MW Lake Mariner NY, Q1 2026 138 MW; Fluidstack; GW/Valuation Opportunity: EV $5.17B on 0.35 GW implies $14.8B/GW, overvalued offset by deal momentum; profitability (low-cost power PUE ~1.4, $1.78M/MW/year) limits 10% downside.
🔹 CORZ (88/100, Crypto Pivot/AI Hosting): 390-590 MW multi, Q1 2026; CoreWeave 250 MW; GW/Valuation Opportunity: EV $6.14B on 0.49 GW implies $12.5B/GW, moderately overvalued; profitability (PUE 1.5-1.6, hosting margins >80%) for 15% upside.
🔹 RIOT (87/100, Crypto Pivot/Hybrid): 600-1 GW Corsicana TX, Q1 2026; site; GW/Valuation Opportunity: EV $7.67B on 0.8 GW implies $9.6B/GW, fair with 20% pivot add; profitability (PUE ~1.5, BTC fallback $0.3-0.4M/MW/year) undervalued on shift.
🔹 GLXY (92/100, Crypto Pivot/Colocation): 800 MW Helios TX, Q4 2025; earnings; GW/Valuation Opportunity: EV $8.41B on 0.8 GW implies $10.5B/GW, slightly overvalued undervalued on narrative; profitability (PUE 1.5, $0.5-0.8M/MW/year) 25% re-rating.
🔹 BITF (85/100, Crypto Pivot/Hybrid): 300-1.3 GW Panther Creek PA, Q1 2026; acquisition; GW/Valuation Opportunity: EV $1.55B on 0.8 GW implies $1.9B/GW, highly undervalued (80% below); profitability (PUE ~1.4, U.S. pivots) strong buy despite risks.
🔹 CLSK (84/100, Crypto Pivot/Hybrid): 300-808 MW GA/TX, 2025 pivot; credit; GW/Valuation Opportunity: EV $4.16B on 0.55 GW implies $7.6B/GW, undervalued (20% below); profitability (clean energy PUE <1.4, $0.3-0.8M/MW/year) 30% upside.
🔹 MARA (82/100, Crypto Pivot/Diversification): 600 MW Granbury TX, Q1 2026; micro-DCs; GW/Valuation Opportunity: EV $9.50B on 0.6 GW implies $15.8B/GW, overvalued on BTC; profitability (PUE 1.5, micro margins) 15% downside unless diversified.
🔹 SLNH (80/100, Crypto Pivot/Hybrid): 83-166 MW Kati TX, Q2 2026; renewable; GW/Valuation Opportunity: EV $0.17B on 0.12 GW implies $1.4B/GW, extremely undervalued (85% below); profitability (renewable PUE <1.4, hosting) niche but scale risks.
🔹 BTDR (78/100, Crypto Pivot/Hybrid): 221 MW Massillon, Q4 2025 1.1 GW; hydro; GW/Valuation Opportunity: EV $3.92B on 0.66 GW implies $5.9B/GW, undervalued (38% below); profitability (hydro PUE ~1.3, colocation) 25% growth.
------------------------------
## 6. Competitive Analysis and Allocation Recommendations
Hyperscalers (70% share, 4.8/5 viability) excel in capex stability but face 20% overruns and PUE 1.1-1.4; crypto pivots (30%, 3.5/5) offer 40% faster timelines with margins >80% yet volatility; niche (30%, 3/5) innovate in fuel cells/PUE <1.2; recommend 70% anchors for IRR, 30% pivots for growth.
🔹 Anchors scale via OpenAI-Broadcom (Reuters 10/13/2025) but cost risks; pivots like GLXY (Helios) 22% premiums, offset volatility.
🔹 Niche underfund (Bloom $5B, DCD 10/13/2025) but sustain at 536 TWh; 99% share, 584.3% CAGRs validate.
🔹 Recommend 70% anchors (MSFT/GOOGL, 15% IRR), 20% pivots (CIFR/IREN, 2-5x), 10% niche (NBIS); Gartner (10/2025) depreciation counter.
------------------------------
## 7. Strategic Playbook and SWOT Insights
Deconstructing power/profitability (50% combined weight), validations link 80% closures to redundancies/low PUE; playbook prioritizes Q4 activations for 2-10x, balancing $400B spend/$12.5M/MW revenue positives with 38 GW strains/$10-20M costs negatives.
🔹 Strengths: Pivots speed 40% (WULF deals); capex drives 33% to 123 GW (McKinsey 10/2025).
🔹 Weaknesses: Delays defer 18% GW, 10-20% IRR hits (JLL 10/2025); 1.6% vacancy limits.
🔹 Opportunities: Fuel cells/liquid cooling cut 50-60% energy (Deloitte 2025); high margins undervalue like BITF.
🔹 Threats: $40B depreciation on 30% efficiency (Gartner 10/2025); mandates inflate 20% (Utility Dive 10/2025).