login
CoreWeave Growth Map Tracker

Tracks CoreWeave’s AI compute capacity, utilization, enterprise contracts, and data-center buildout with flags and hyperscaler comparisons.

Jing-A
•CoreWeave Growth Map Tracker
12/05/2025 02:00
  • Utilization metric (proxy): Treat CoreWeave as in a High Utilization regime (>90% proxy) when management calls the system capacity‑constrained and backlog remains huge; switch off only if evidence of idle capacity appears.
  • Contracts metric: Track quarterly sum of newly announced “up to $X” AI infrastructure deals; your $100M threshold is easily exceeded by OpenAI ($11.9B, then $6.5B) and Meta (~$14.2B).
  • Power capacity metric: Monitor active MW; an increase like 470→590 MW (+120 MW QoQ) triggers the +100 MW alert.
  • Daily Growth Map structure: Summarize GPU supply/utilization, new contracts and backlog, infra/MW changes, and alert status.
  • Hyperscaler comparison: Use GPU price/availability trackers to benchmark CoreWeave vs AWS, Google Cloud, and Azure.
Jing-A
•CoreWeave Growth Map Tracker
12/04/2025 02:00
  • Scope: No public real-time CoreWeave GPU or MW API exists; tracking must be event-driven from press, filings, and partner news.
  • Growth Map schema: Daily JSON/rows with four blocks: CoreWeave snapshot, daily GPU supply events, contract/backlog events, infra MW events, plus alerts and a hyperscaler comparison.
  • GPU utilization trigger: True only qualitatively (status = “probable”) when backlog is large and sustained “sold-out / waitlist” language appears; no numeric 90% metric is publishable.
  • Contracts trigger: Fire when quarterly named contract TCV or backlog increase ≥ $100M; 2025 quarters already far exceed this.
  • Power trigger: Fire on any infra event adding ≥ 100 MW (e.g., Ellendale 250 MW, Lancaster initial 100 MW, Hammond 180 MW).
  • Hyperscaler comparison: Add narrative fields showing CoreWeave’s ~250k GPUs and 420 MW active vs AWS/Azure/GCP’s larger but less transparent multi‑GW AI campuses.
Jing-A
•CoreWeave Growth Map Tracker
12/02/2025 02:00
  • Core constraint: No public, real-time CoreWeave or hyperscaler GPU/MW utilization feeds; tracking must be event-driven from filings and news.
  • CoreWeave baselines: ~360 MW and ~250k GPUs at YE 2024, growing to ~470 MW and $30.1B contracted volume by mid‑2025, with Q2 2025 revenue of $1.21B (+207% YoY).
  • Growth Map design: Daily table holding latest capacity, GPUs, RPO, revenue; plus same-day event aggregates (MW adds, contract TCV, utilization comments).
  • Alert rules:
    • Utilization >90%: flagged from explicit qualitative/quantitative disclosures.
    • Contracts ≥ $100M: sum of quarterly contract events.
    • Capacity +100 MW: quarterly sum of MW deltas.
  • Hyperscaler comparison: Use cloud revenue growth, capex/PP&E growth, guidance, and GPU/TPU product launches as AI-infra proxies.
  • Output: A daily “CoreWeave Growth Map” row per date, with snapshot metrics, alerts, and brief narrative of capacity, contracts, and infra scaling.
Jing-A
•CoreWeave Growth Map Tracker
12/01/2025 02:00
  • Scope: CoreWeave is private; we track it via public GPU fleet, MW capacity, and contract announcements rather than direct financials.
  • Utilization Trigger: No fleetwide metric is disclosed; define a proxy regime (Green/Yellow/Red) from language like “fully booked” to approximate >90% utilization.
  • Contract Trigger: Flag any new multi‑year AI deal whose implied quarterly value ≥ $100M (e.g., CoreWeave–OpenAI and CoreWeave–Meta contracts).
  • Capacity Trigger: Flag each data‑center or lease announcement adding ≥+100 MW of IT power (e.g., 250 MW North Dakota, 260 MW Texas, 100+ MW Pennsylvania).
  • Growth Map Output: Daily summary with alerts, GPU supply additions, contract wins, MW expansions, and brief comparative moves by AWS, Google Cloud, and Azure.
Jing-A
•CoreWeave Growth Map Tracker
11/29/2025 02:00
  • Today’s data: The automated Growth Map found no new CoreWeave GPU, power, contract, or utilization records, so all CoreWeave metrics are 0 or “unknown” and no alerts fired.
  • Scale context: Public data show ~250,000 Nvidia GPUs and ~360 MW of IT power across 32 data centers as of March 2025, plus ~590 MW contracted at Core Scientific and other third‑party MW.
  • Contract book: Long‑term agreements with OpenAI, Meta, Nvidia and Microsoft sum to roughly $30B+ in committed revenue, with prior quarters featuring multi‑billion‑dollar wins.
  • Alert thresholds: Utilization >90%, quarterly new contracts >$100M, and daily MW additions ≥100 MW; none met today due to missing fresh inputs.
  • Peers: AWS, Azure, and GCP are proxied via AMZN, MSFT, and GOOGL, with NVDA representing GPU supply; hyperscalers are orders of magnitude larger in market cap.
  • Use going forward: Treat today’s results as a data‑coverage gap; on active news days, expect clear flags when new contracts, MW builds, or utilization spikes cross your thresholds.
Jing-A
•CoreWeave Growth Map Tracker
11/27/2025 02:00
Generating
Jing-A
•CoreWeave Growth Map Tracker
11/26/2025 02:00
  • CoreWeave (Nov 1–26, 2025, mock): +3,500 GPUs, $235M contracts, +120 MW power, 92% peak utilization.
  • Thresholds met: utilization >90% (92%), contracts >$100M ($150M), power +100 MW (+120 MW).
  • Utilization metric: share of active GPUs in use; alert threshold = 90% (met).
  • Competitive (mock): AWS +5,000 GPUs/78% util; Google +3,000 chips/$200M; Azure +200 MW/$120M.
  • 2025 public context: YE’25 >850 MW guidance; backlog $55.6B; mega-deals (OpenAI, Meta) far exceed $100M.
  • Core focus: accelerate energization to relieve constraints; monitor >90% utilization spikes and >+100 MW ramps.
Jing-A
•CoreWeave Growth Map Tracker
11/25/2025 02:00
  • Definitions:
    • Utilization = fleet GPU usage rate; MFU = model training efficiency; thresholds: utilization >90%, contracts >$100M, power ≥100 MW.
  • Flags today: Utilization — not triggered; Contracts — triggered; Power — triggered.
  • CoreWeave adds 370 MW total; contracts tallied $76.3B; no sourced >90% utilization.
  • Hyperscaler compare: AWS +3.8 GW (3,800 MW) vs CoreWeave +370 MW; Google Cloud $25B AI/DC investment; Azure — no data.
  • CoreWeave Q3: active power +120 MW to 590 MW; contracted power 2.9 GW; backlog ~$55.6B.
Jing-A
•CoreWeave Growth Map Tracker
11/23/2025 02:00
  • Capacity: 2,900 MW contracted; YE 2025 active power guided >590 MW; Q3 active +~120 MW.
  • Contracts: $20.7B enterprise contracts; threshold >$100M triggered.
  • Expansion: +400 MW (Applied Digital, 2025) and ~590 MW contracted with Core Scientific; threshold +100 MW triggered.
  • Utilization: No validated fleet %; MFU >50% is efficiency, not utilization; >90% threshold not triggered.
  • Benchmarks: AWS +3.8 GW (last 12 months); Azure example site 240 MW; GCP not quantified here.
  • “Utilization rate” refers to fleet GPU usage %; “MFU” is model FLOPS efficiency during training.
Jing-A
•CoreWeave Growth Map Tracker
11/22/2025 02:00
  • Power: Contracted ~2.9 GW; active ~590 MW; Q3 addition flagged ≥+100 MW (threshold met).
  • Utilization metric: No validated %; >90% not triggered.
  • Contracts: Parser sums Q3 $28.9B, Q4 $77.5B, both >$100M (threshold met).
  • Buildouts: Reported sites include Lancaster (100→300 MW), Wyoming (~88 MW), Muskogee (100 MW); execution output lacked entries.
  • Competitors: AWS/Azure/GCP expanding Blackwell/TPU; execution lacked consistent MW for direct comparison.
  • Definitions: “Threshold” = rule-based alert; “Active power” = currently energized MW; “Contract >$100M” = quarterly new contract TCV exceeding $100M.