CORZ
Core Scientific
Summary
What they do:
Core Scientific operates data center campuses across Texas, Georgia, North Carolina, Oklahoma, and Alabama, converting former Bitcoin mining sites into AI/HPC hosting facilities under long-term contracts with CoreWeave and other customers.
Why they matter:
CORZ holds the largest single AI hosting contract in the L24 layer — $10B+ with CoreWeave across ~590MW — and controls 1.3GW of contracted power, making it the most scaled miner-to-AI pivot in public markets.
Recent performance:
Q4 2025 revenue $79.8M with net income of $216M (accounting gain); colocation revenue $31.3M (+268% YoY); stock ~$20, market cap ~$6.3B. Revenue declined 38% for full year 2025 as Bitcoin mining wound down.
Our Verdict
The largest miner-to-AI-host conversion with $10B+ in CoreWeave contracts and 1.3GW of power, trading at ~0.6x contracted backlog — execution on delivery milestones and customer diversification beyond CoreWeave determine whether this becomes a $15B infrastructure platform or stays a single-customer dependency.
Structural trends
Structural
65
/ 100
Moat
5/10
Scaled Power + CoreWeave Lock-in
AI Exp.AI Exposure
High~40% AI
Play Type
EmergingAI Growth
~268% YoY
Rel. Value
57
ATTRACTIVEPriceLIVE
$18.68
-0.16%
Live via Yahoo Finance · refreshes every 5 min
Market Cap
$5.9B
P/E Ratio
N/A
P/S Ratio
18.5x
52W High
$23.63
52W Low
$6.30
52W Chg
196.5%
Beta
6.36
Core Scientific emerged from Chapter 11 bankruptcy in January 2024 with a clean balance sheet and an asset base that turned out to be worth far more for AI than for Bitcoin mining. The company operates six data center campuses across Texas (Denton, Austin, Pecos), Georgia, North Carolina, and a new facility in Muskogee, Oklahoma. These sites collectively represent approximately 1.3GW of contracted power — the raw material for AI infrastructure.
The pivot from Bitcoin mining to AI hosting has been dramatic. In Q4 2024, colocation revenue was $8.5M; by Q4 2025, it reached $31.3M. Bitcoin mining revenue went the other direction — from $79.9M to $42.2M in the same period. The company is actively converting remaining mining capacity to HPC hosting, with management stating that 2026 priorities include "diversifying the customer base and executing on the CoreWeave contract milestone delivery."
The CoreWeave relationship dominates the business. Through a series of contract options exercised in 2024-2025, CoreWeave committed to approximately 590MW of critical IT load across six Core Scientific sites, with cumulative contracted revenue exceeding $10B over 12 years. The Denton, Texas expansion alone added approximately 100MW of gross capacity (~70MW critical load) through a $1.2B agreement, with operational target of mid-2026. About 350MW is energized, with over 180MW online and billing.
CoreWeave attempted to acquire Core Scientific outright in July 2025 for approximately $9B, but the merger was terminated in October 2025 after shareholders voted it down, believing the company's standalone value exceeded the offer. The commercial relationship survived intact — CoreWeave remains the anchor tenant, and the failed merger arguably strengthened CORZ's negotiating position.
New facilities signal diversification. The Muskogee, Oklahoma campus (Phase 1: 70MW) reached full energization and entered commissioning in early 2026, on track for Q2 delivery. An Auburn, Alabama facility — a smaller, multi-tenant, Tier 3 data center — is under construction with the first 10MW expected in H2 2026, designed specifically for enterprise AI and inference workloads to serve customers beyond CoreWeave.
Supply Chain Dependencies
Upstream Suppliers
Downstream Customers
The Catch
Core Scientific's existential risk is customer concentration. CoreWeave accounts for 70-75% of contracted future revenue, and CoreWeave itself is a young, unprofitable company with $1.17B in 2025 net losses and massive debt. If CoreWeave faces financial stress — a GPU pricing shift, a loss of its own major customers, or a capital markets disruption — CORZ's $10B+ backlog could be impaired. The accounting restatement adds a secondary risk: a post-bankruptcy company with material control weaknesses inspires less confidence in financial reporting. These are not theoretical risks — they are the specific reasons CORZ trades at a 30%+ discount to its own rejected acquisition price.
If They Win
If Core Scientific successfully delivers the full 590MW CoreWeave pipeline on schedule, diversifies with 2-3 additional investment-grade tenants using its 1.3GW power base, and achieves the 55-65% EBITDA margins management targets, it becomes the definitive infrastructure backbone for the AI compute buildout. In this scenario, CORZ generates $1.5B+ in annual revenue with $800M+ EBITDA, supports a $15-20B enterprise value, and potentially converts to a REIT structure — creating a new category of AI infrastructure income stock. The failed CoreWeave acquisition becomes a footnote: the company shareholders rejected a $9B offer for what turned into a $20B business.
Others in Operate the Data Center
Not financial advice. All scores generated via AI algorithms using public data.