Skip to content

CORZ

Core Scientific, Inc.

NASDAQ · Technology · Software - Infrastructure · US

$17.89
−0.06%
Ask drillr

Research · Sep 3, 2026

[CORZ] Core Scientific Thesis 2026: CoreWeave HPC Hosting Contracts Reshape the Bitcoin Miner

Core Scientific, Inc. (NASDAQ: CORZ) is a US digital infrastructure company — Bitcoin self-mining and hosting transitioning toward HPC/AI data center colocation hosting — that re-listed on the NASDAQ on its January 2024 emergence from Chapter 11 bankruptcy. CORZ enters FY2026 with FY2025 revenue ~$0.5-0.8B and a fast-shifting mix: ~$0.35-0.55B aggregate Digital Asset Self-Mining + Hosting (Bitcoin) revenue plus ~$0.10-0.30B aggregate HPC/AI Data Center Hosting (CoreWeave) revenue that is ramping toward the dominant share mid-decade, with adj. EPS highly variable (~-$0.50 to +$0.50) on Bitcoin price, HPC ramp timing and non-cash marks, all under CEO Adam Sullivan (CEO since ~2023, ~2-3 year tenure, investment banking / restructuring background, architect of the Chapter 11 emergence and the HPC/AI pivot). The first thesis pillar is the HPC/AI Data Center Hosting (CoreWeave Colocation) pipeline: a ~12-year ~$8.7-10B+ aggregate contracted hosting backlog with CoreWeave (CRWV), under which Core Scientific is converting and expanding sites (Denton Texas and others) to deliver ~590MW+ aggregate critical IT load for CoreWeave's GPU compute fleet via a powered-shell, critical-infrastructure colocation model — CoreWeave supplies the GPUs/servers, Core Scientific supplies power, datacenter shell, cooling and operations — generating stable contracted cash flows in contrast to volatile Bitcoin mining, with expansion options for additional MW and an additional HPC/AI customer pipeline; the model requires ~$1-3B+ aggregate FY2025-FY2027 conversion capex (a mix of convertible notes, project-level financing and CoreWeave-related financing) with milestone-based revenue commencement, and FY2026 catalyst is ~$0.4-1.0B+ HPC/AI hosting revenue at ~70-85% gross margin as more CoreWeave MW come online. The second pillar is the Digital Asset Self-Mining + Hosting (Bitcoin) pipeline (~$0.35-0.55B revenue, declining mix): owned-and-operated Bitcoin mining fleet with exahash capacity post the April 2024 halving, fleet efficiency upgrades, legacy third-party hosting contracts (some being converted to self-mining or HPC), and — critically — the ~1,200MW+ aggregate power asset base across six states with low-cost power purchase agreements and grid interconnection, the strategic asset whose MW are far more valuable converted from Bitcoin mining to HPC hosting; FY2026 catalyst is ~$0.3-0.6B Digital Asset revenue with capital allocation increasingly toward HPC conversion. The capital story: no dividend, no buyback (capital reinvestment, HPC buildout priority), ~$0.5-2.0B net debt (convertible notes plus HPC project financing on a post-Chapter 11 fresh-start balance sheet, with a contingent-value-rights overhang resolving), highly variable net leverage near-term (capex-heavy buildout), a non-rated to B-/B3 credit profile, ~290-320M diluted shares (warrants and RSU dilution), plus some bitcoin holdings with mark-to-market exposure; near-term net debt rises on the buildout, then de-levers as contracted HPC cash flows commence. At ~$8-25 per share on ~290-320M shares (~$3-8B equity, ~$4-10B EV) CORZ trades at ~8-20x EV/EBITDA and ~5-15x EV/Sales on ramping mid-decade contracted revenue versus Bitcoin-miners-turned-HPC-hosts Iren/Iris Energy, Cipher Mining, TeraWulf, Hut 8, Bitdeer, Applied Digital, Riot Platforms and MARA Holdings, with traditional data center operators Equinix and Digital Realty and the customer CoreWeave as HPC comps. FY2026 base case is ~$0.8-1.5B revenue + ~$0.3-0.7B adj. EBITDA with HPC ramping toward the dominant mix and elevated buildout-related net debt; bull case ~$1.5-2.5B+ revenue + ~$0.7-1.3B+ adj. EBITDA on faster CoreWeave MW delivery, expansion-option exercises, additional HPC customers and a de-risking re-rating; bear case ~$0.5-0.8B revenue + ~$0.1-0.3B adj. EBITDA on CoreWeave counterparty-concentration and credit risk (CoreWeave's funding, GPU demand and AI capex cycle is the key dependency), competitive intensification from Iren, Cipher, TeraWulf, Hut 8 and Applied Digital, AI/GPU demand-cycle risk, site-conversion execution and cost overruns, buildout financing risk, Bitcoin price volatility and the post-2024 halving headwind, power-cost and grid-interconnection issues, capital-allocation tension (mining vs HPC) and fresh-start dilution/CVR considerations. The thesis depends on the HPC/AI Data Center Hosting (CoreWeave Colocation) pipeline plus the Digital Asset Self-Mining + Hosting (Bitcoin) pipeline plus the ~$10B+ ~12-year contracted backlog plus the ~1,200MW+ power asset base plus the post-Chapter 11 fresh-start balance sheet and Adam Sullivan's execution of the CoreWeave contracts and site-conversion capex.

Research · Apr 23, 2026

Can Core Scientific's $3.3 Billion Junk Bond Actually Fund the AI Pivot?

Core Scientific's $3.3 billion junk bond offering removes capital availability as a constraint on its AI infrastructure pivot, but the real alpha emerges when Q2 and Q3 filings reveal whether 70%+ of proceeds fund capacity expansion versus debt repayment. Current valuation at $13.47 embeds a 68% discount to the bullish scenario, creating asymmetric upside if allocation skews to AI buildout.