META — Compute Capacity Build-Out to 2030

First-principles look at what Meta is actually building, owning, leasing, and contracting — and whether the ~$1.7T market cap reflects it. Updated 2026-05-19. v2 (corrects v1's errors on Hyperion ownership and 2025/2026 capex).
Correction notice (v2). v1 of this report stated Hyperion was "Meta-owned" and that Meta owns/self-builds ~85–90% of its 2030 capacity. That was wrong. In October 2025 Meta sold 80% of the Hyperion JV to Blue Owl-managed funds and converted to an operating-lease structure on the campus[1][2]. v1 also used stale capex numbers; Meta's 2025 actual was $72.22B and 2026 guidance is $115–135B[3]. Both are fixed below. The qualitative thesis still holds but the magnitudes and the "what Meta owns" picture are materially different from what I wrote in v1.
Bottom line up front. By end of 2030, Meta is on track for roughly 10–14 GW of operational AI-capable IT load. But the ownership structure is far less clean than I originally wrote: Hyperion alone (5 GW at full build) is sitting in an 80/20 JV with Blue Owl with Meta as the tenant on initial 4-year operating leases[1]. Realistic economic interest in the underlying real estate is closer to 55–65% owned (the legacy fleet) and 35–45% structured/leased (Hyperion JV + likely future SPVs). GPUs are still on Meta's balance sheet. Replacement value all-in (shell + GPUs + power infra) by 2030 is plausibly $700B–$1.1T; Meta's economic share of that is closer to $500–800B. Against a $1.7T market cap, the compute is still under-priced versus an AGI-strategic-asset framing — but the JV-lease financing structure means a chunk of the upside leaks to Blue Owl / PIMCO bondholders.

1. Total compute capacity by end of 2030

ScenarioIT load (GW)B200-equivalent GPUsNotes
Bear~8 GW~5–6MHyperion phases slip; one major campus delayed 2+ years; permitting/grid issues
Base~11 GW~7–8MHyperion hits ~2 GW by 2027, ~5 GW by 2030; Prometheus full; fleet of mid-size campuses online
Bull~14 GW~9–10MHyperion full 5 GW by 2030; second mega-campus announced and partially online; MTIA displaces a large share of Nvidia spend

Zuckerberg's Jan 2025 Facebook post explicitly framed the build as Hyperion reaching ~5 GW eventually, with footprint "covering a significant part of the footprint of Manhattan"[4]. The Blue Owl JV press release confirms ~5 GW at full build for Hyperion[1]. Combined with the existing 25+ campus fleet, base-case 2030 lands at 10–12 GW.

Reference points: Meta's installed base going into 2025 was ~5 GW across the global fleet (estimates from SemiAnalysis and DCD)[5]. Google's 2030 path is comparable (10–15 GW), Microsoft is in a similar range largely via leases, Amazon is the biggest but most fragmented. Meta is in the top tier — not Stargate scale (5+ GW from a single project), but in the small club of operators that will own/operate multi-GW AI campuses.

2. What Meta owns vs. leases vs. structured-finance

2a. The Hyperion JV (the big correction)

The Oct 27, 2025 announcement is structurally important and was the load-bearing error in v1[1][2][6]:

Economic interpretation: This is the same financing pattern Microsoft uses for some of its CoreWeave-adjacent capacity and that Oracle uses with Crusoe — get the campus built using third-party capital, pay rent rather than capex, keep more cash for GPUs and OpEx. The difference is Meta has a 20% equity strip and an RVG, so it retains meaningful upside if the campus is worth a lot in 2042. But the headline "Meta owns Hyperion" is wrong — Meta is the 20% LP and the tenant.

Why Meta did this: the ~$3B distribution + the ability to expense lease payments rather than depreciate ~$27B of capex over 20+ years materially improves near-term reported earnings and frees cash for the bigger ticket item — GPUs. Wall Street's reaction was positive precisely because it offloaded capex intensity at a time when 2026 capex was about to be guided to $115–135B[3].

2b. Owned / self-built fleet (the legacy estate)

CampusStatusCapacityOwnership
Prometheus — New Albany, OHPhasing online 2025–26~1+ GWMeta-owned
Hyperion — Richland Parish, LAUnder construction~5 GW at full build20% Meta / 80% Blue Owl JV — Meta leases[1]
Existing fleet (~25 owned campuses)Operational~5 GW today, growingMeta-owned
Next mega-campus(es)Site-hunting / partially announced2–5 GW eachLikely structured similarly to Hyperion JV given the financing precedent

Existing fleet (operational, owned): Forest City NC, Prineville OR, Altoona IA, Luleå Sweden, Clonee Ireland, Henrico VA, Newton GA, Los Lunas NM, Eagle Mountain UT, Papillion NE, Sarpy NE, Fort Worth TX, Temple TX, DeKalb IL, Huntsville AL, Kansas City MO, Cheyenne WY, Gallatin TN, Rosemount MN, Jeffersonville IN, Montgomery AL, Stanton TX, Idaho Falls ID, Beaver Dam WI[7].

2c. Pure leases and colocation

Historically small. Meta has used some build-to-suit deals (Digital Realty, CyrusOne, QTS in earlier years) but third-party colo is a small minority of its footprint. With the Hyperion JV, the operating-lease share of the balance sheet just jumped materially.

2d. Neocloud contracts

Meta is still an outlier here vs. peers. Microsoft has multi-tens-of-billions committed to CoreWeave; Oracle is leaning heavily on Crusoe; Google has selectively contracted. Meta has done little of this publicly at scale. The Blue Owl JV is the inverse of a neocloud contract — Meta uses third-party capital but operates the asset itself.

Updated economic-ownership split (rough, base case 2030, ~11 GW): Equity-weighted economic share of physical real estate / shell: roughly 6 GW + 1 GW = ~7 of 11 GW = ~65%. Compared to v1's "85–90% owned" — meaningfully lower.

3. Capex trajectory (corrected)

YearCapexSource
2023$28.1B10-K
2024$39.2B10-K
2025 (actual)$72.22B (incl. principal on finance leases)Q4 2025 release[3]
2026 (guided)$115–135B (incl. principal on finance leases)Q4 2025 release[3]
2027Plausibly $140–170BTrajectory — Meta has flagged "notably larger" growth and Superintelligence Labs build-out[8]
2028–2030Likely plateau in $130–170B/yr bandHyperion 5GW build alone is ~$27B over its lifecycle; new sites likely structured JV-style
Cumulative 2025–2030~$750B–$1.0TImplied by trajectory above — meaningfully higher than v1's $550–750B estimate

Important: this is the largest single-company capex program in history — comparable in real-dollar terms to the Apollo program, the US interstate system, or peak-era AT&T. It is funded by a combination of (a) Family of Apps operating cash flow, which still runs $90–110B/yr in EBITDA, (b) the Hyperion-style JV structures that move shell capex off-balance-sheet, and (c) a modest increase in debt issuance (Meta floated $30B of bonds in late 2025).

4. The stack — what Meta actually owns

LayerOwned?Detail
Land (legacy fleet)Yes (freehold)Tens of thousands of acres aggregated across ~25 campuses
Land (Hyperion)JV-owned, Meta 20%~2,000+ acres in Richland Parish, LA — owned by Blue Owl-led JV
Buildings & mechanical (legacy)Yes (self-designed)Open Compute reference design, modular
Buildings & mechanical (Hyperion)JV-owned, Meta tenantMeta provides construction management services to the JV
Power infrastructure (legacy)Mostly yesOwns substations and backup gen on the legacy campuses
Power infrastructure (Hyperion)JV-ownedIncluding the long-lived power/cooling/connectivity infra — explicitly enumerated as JV assets[1]
Energy contracts (PPAs)Meta-contracted30+ GW of clean-energy PPAs cumulatively. Constellation Clinton nuclear restart ~1.1 GW, 20-year[9]. Multiple SMR developer agreements signed 2024–2025
GPUs (Nvidia)Owned by Meta~600K H100-equivalents end of 2024; targeting 1.3M GPU-equivalents end of 2025 per Zuckerberg[4]; trajectory to 5–8M B200-equivalents by 2030. These sit on Meta's balance sheet even when housed in the Hyperion JV's shell
Custom silicon (MTIA)Owned IP, fab via TSMCMTIA v2 ramped 2024–25 for ranking/recommendation inference. v3 in development for training. Plausibly handles 30–50% of inference workload by 2028, compressing the Nvidia bill
NetworkingOwnedCustom in-house DC switch fabric, optical interconnect
CoolingOwned (legacy) / JV (Hyperion)Shifting to direct-to-chip liquid for AI racks
Models & dataOwnedLlama family, training data corpus, RLHF infra

Key reframing: Meta still owns the value-dense piece (the GPUs) and the optionality piece (the models, data, MTIA IP). What it has partially sold off is the lowest-IRR, highest-capital-intensity piece (shell + power infra at Hyperion). That's actually rational capital allocation if you believe GPUs and IP carry the strategic premium and shells are a real-estate cost.

5. Replacement value (updated)

Sanity-check the asset on a rebuild-from-scratch basis. Industry rule of thumb for greenfield AI DC construction in 2025–2026[5][6]:

2030 capacityGross replacement (shell)Gross replacement (all-in)Meta equity-weighted share (all-in)
8 GW (bear)$320–480B$600–880B$430–630B
11 GW (base)$440–660B$825B–$1.21T$575–845B
14 GW (bull)$560–840B$1.05–1.54T$735B–$1.08T

The "Meta equity-weighted" column applies a haircut: for the ~6 GW Meta fully owns, count 100%; for the ~5 GW in Hyperion-style JVs, count GPUs (~100%, since GPUs aren't in the JV) plus 20% of the shell. That gives roughly 70% of gross all-in value as Meta's economic share, which is the column shown.

What this is and isn't. Replacement value is not market value. The binding constraints are power interconnects (Constellation/Entergy interconnects took years), land in approved jurisdictions, and GPU allocation. The functional value to someone who needs compute right now is materially higher than rebuild cost. But for Meta-the-equity-holder, the JV structure means some of that scarcity premium accrues to Blue Owl, not Meta.

6. What is the compute worth, really?

Lens A — Replacement cost (Meta's economic share)

~$575–845B in 2030, base case. Conservative but a defensible lower bound.

Lens B — Neocloud comparable

CoreWeave at peak traded 8–15x forward revenue on contracted AI capacity. At ~$25M/MW-yr fully utilized rental revenue for B200-class capacity, 11 GW = $275B/yr theoretical revenue. Apply a 10x multiple: $2.75T gross. Even with a steep haircut for Meta consuming most of its own capacity (so no actual revenue) and the Hyperion JV split, the opportunity-cost framing says Meta's economic interest in the compute fleet is worth well over a trillion if compute itself stays scarce.

Lens C — Strategic value under AGI premise

If recursive self-improvement starts in 2026–2027 and the binding constraint on intelligence becomes compute, then 7+ GW of Meta-controlled fleet by 2028–2030 is not a depreciating asset — it's a seat at the table. Likely 5 entities globally with this scale: Google, Microsoft, Amazon, Meta, and the OpenAI/Stargate complex. Anyone training a frontier model must negotiate with one of them or wait years.

Under that lens the right comparable isn't EQIX/DLR (20x EBITDA) or even CoreWeave. It's "essential strategic infrastructure during a phase change" — Standard Oil's pipeline network, AT&T's long-distance lines pre-divestiture, TSMC's leading-edge fabs today. Such assets at peak strategic importance don't carry a normal multiple — they're priced on optionality and bargaining power.

7. Mapping back to the $1.7T market cap

ComponentImplied valueComment
Family of Apps (ads business)~$1.2–1.4T at 18–22x earningsThe market essentially prices Meta as ads + a low residual for AI
Cash & investments (net of debt)~$30–50BDown from prior years after $30B bond issuance + Hyperion-related $3B inflow
Reality LabsNet negative ($15B+/yr drag)Market gives it zero or negative value
Implied value of Meta's compute infrastructure equity~$250–500BThe residual after ads + cash – RL drag. Below Lens A replacement cost
So is it priced in? Still no — but the gap is narrower than v1 suggested. Even on Lens A (replacement cost of Meta's equity share), the market is paying ~$250–500B for an asset that will cost Meta itself $575–845B to build over 2025–2030. The market is treating capex as expense and assuming much of the compute will be obsolete by the time it's online. If you believe AGI is happening, this is still one of the largest mispricings in mega-cap equity — just call it a $300–500B mispricing rather than a $700B+ one.

8. The honest pushback

  1. Llama is behind frontier. Meta's models are roughly Gemini 2.5-tier as of mid-2026 — 12–18 months behind GPT/Claude/Gemini-frontier. If recursive self-improvement starts at the leading lab and pulls away, owning compute without a competitive model matters less. Counter: compute is rentable, partnerable, or sellable to whoever wins; the asset isn't model-specific.
  2. GPU depreciation is brutal. A B200 in 2030 may be worth 20% of original cost. The $300–450B GPU portion of replacement value turns over twice during 2025–2030. The compounding capex is real and is exactly why Meta is using JV financing on the shells.
  3. Power constraints may slip Hyperion. Entergy's grid build and the gas turbine permits are not done. 1–2 year slip on Phase 2 is plausible → base case drops to 8–9 GW.
  4. The JV structure dilutes the win. If compute becomes radically more valuable than 2025 underwriting assumes, Blue Owl/PIMCO captures most of that on the Hyperion shell. Meta keeps 20% equity + the RVG mechanism. That's a real leakage that wasn't in v1's math.

None of these change the qualitative thesis: Meta is one of 4–5 globally significant compute operators of the AGI era and the market is paying for it as if compute is a cost line. They change the magnitude — the right framing is "compute infrastructure adds $300–500B to fair value above the standalone ads business" rather than v1's "$1T+."

9. What to watch

Sources

  1. Meta investor release, "Meta Announces Joint Venture with Funds Managed by Blue Owl Capital to Develop Hyperion Data Center," Oct 27, 2025. investor.atmeta.com
  2. Datacenter Dynamics, "Meta forms $27bn joint venture with Blue Owl to fund gigawatt-scale AI data center campus in Louisiana." datacenterdynamics.com
  3. Meta, "Meta Reports Fourth Quarter and Full Year 2025 Results," Jan 28 2026. 2025 capex $72.22B; 2026 guidance $115–135B. investor.atmeta.com
  4. Mark Zuckerberg Facebook post, Jan 24 2025 — Hyperion ~5 GW, Prometheus >1 GW, ~1.3M GPU-equivalents target. facebook.com/zuck
  5. SemiAnalysis and DCD reporting on hyperscaler GW footprint, GPU shipment ranges, and per-MW cost stack (multiple posts 2024–2026). datacenterdynamics.com analysis
  6. Global Data Center Hub, "Meta + Blue Owl's $27B Bet: Is This the US Blueprint for Financing AI Data Centers?" globaldatacenterhub.com
  7. Meta data center locations directory. datacenters.atmeta.com
  8. Datacenter Dynamics, "Meta plans 'notably larger' capex spend on AI data centers in 2026." datacenterdynamics.com
  9. Constellation Energy press release on Crane (formerly TMI) restart with Meta-style PPA precedent — and Meta's own Constellation Clinton nuclear PPA announcement (June 2025). constellationenergy.com
Some forward numbers (B200 equivalents, MTIA share of inference, replacement cost per MW) are first-principles estimates synthesized from SemiAnalysis ranges and sell-side commentary, flagged in text where so. The Hyperion JV economic interpretation (20% equity + tenant + RVG ≠ "owned") is from a direct read of the Oct 27, 2025 Meta release[1]. v1 of this report (now superseded) is preserved in Google Drive under the same folder.