Space digest — 2026-10-10
Top line: A financing-and-enablement window: Blue Origin put an IPO ("within the next few years") on the table along with a $550M Texas satellite factory, SpaceX locked in the Cape's fuel supply and a 15,000-satellite carrier clearance, and NASA signed a nuclear MOU with DOE — while Starship's Florida flights slipped to NET Nov 8/Dec 10 and Amazon's 1,000th satellite exposed the not-enough-rockets bottleneck.
1. Blue Origin: IPO signal, Constellation Park, and the LC-36 ground game
What happened: Bezos said for the first time Blue Origin "could go public within the next few years" ("it probably makes sense" — Fox interview, multi-outlet). One day prior (Oct 8): Constellation Park — a $550M, 1.3M-sq-ft satellite factory in Texas (~2,000 jobs) to build the TeraWave constellation — as Bezos also confirmed the $10B equity round. Ground side (LC-36A): the launch bridge is "back and ready for use" — structure salvaged, bearing surfaces reworked, main water piping reinstalled (Limp, Oct 9) — and the New Glenn wreckage was cleared in just 9 days.
What's genuinely new: the IPO is the first public filing-track signal in the company's history, one week after the $10B round's fine print (~$28B of Bezos money in, $140B valuation, ~$800M 2025 revenue). The factory is the concrete instrument behind the revenue mix: TeraWave was recorded "yet to fly a satellite" in the Sep 26 round entry — now with a named 2,000-job Texas fab. The May disaster context ($150M vehicle) frames the recovery speed. RTF watch unchanged (ledger due Dec 31).
Watch: what the IPO chatter is for (the next raise), TeraWave's first satellite, RTF, and the module-stacking pace ("four more to go").
Links: Fox/coverage (via feed cluster; Bezos interview) · Blue Origin press (Constellation Park, Oct 8)
2. Starship: the Cape's fuel closes, flight dates slip, and Ship 40 comes home
What happened: (1) The fuel problem closes: Florida regulators approved the Coastal Connect tariff and SpaceX's subsidiary filed for a 32.4-mile gas pipeline through Brevard County to feed Starship — sized for up to 76 launches/yr (SLC-37/39A; the "truck deliveries can't keep up" note is the why). FAA input window open on the Starship-Florida plans; four Cape environmental reviews advanced (Starship, New Glenn, Terran R). (2) F15/F16 slipped (unofficial schedule, Foust): NET Nov 8 (Pad 2 catch) / Dec 10 (39A debut) vs planning dates of Oct 19/30 a week prior; Pad 2 chopsticks exercised for the catch attempt. (3) Ship 40 home: the Flight 13 recovery arrived at Port Brownsville Oct 7 — ~75 days / ~12,000 nm at sea ("Odysseus's voyage home took ten years. Ship 40 took 75 days") — as Ship 42 went to Massey's for cryo proof tests and McGregor ran a 557-second Raptor firing (second-longest ever) plus F9-booster static fire.
What's genuinely new: the pipeline/tariff was the thread's Oct 5 docket watch — now filed (76/yr sizing makes it structural, not incremental); the slip is a checkable date-set (Foust's comparison vs the prior week's planning is the record); and the ocean-recovery return is the first full-cycle data point on reusability logistics.
Watch: F15 actually flying in November, the pipeline docket, pad-order confirmation at 39A, whether F16's 39A debut holds Dec 10.
Links: NSF Starship updates · Foust via feed (slip)
3. SpaceX as a US carrier: FCC clears 15,000 satellites and the no-carrier waiver
What happened: FCC approved the Starlink Mobile application Oct 6 — up to 15,000 next-gen satellites for direct-to-device service (pitched at "150 Mbps to phones") — plus a waiver allowing SpaceX to offer the service without going through a traditional wireless carrier. Spectrum leg: an ~$8B (reported) acquisition of 800 MHz low-band — "the last critical piece of the spectrum puzzle" (Musk; the figure is single-explainer grade). Rival layer: AST SpaceMobile sank ~7–8%, towers wobbled, and the FCC votes Oct 29 on a D2D spectrum auction that opens the lane to rivals; Musk's V2 "100× bandwidth" claim rides along.
What's genuinely new: the carrier-of-record crossing — the waiver (not the constellation) is what changes the business model; everything else in the window is the price discovery (the $40B-debt report, below, is the financing side of the same turn). The India fight escalates in name (Musk naming Ambani vs the security-clearance/spectrum-pricing counterpoint) — recorded in the thread entry.
Watch: the Oct 29 auction vote, waiver challenges, commercial-service start, V2 bandwidth at second source.
Links: FCC approval (feed cluster, multi-outlet) · Explainer (spectrum, via feed)
4. Nuclear NASA: Isaacman signs the DOE MOU
What happened: Administrator Isaacman (X, Oct 8): "We're getting America underway on nuclear power in space and ushering in the era of a Nuclear NASA. With the signing of an MOU with @ENERGY, we're taking the next giant leap toward…" (snippet cut). Companion in-stream: the Artemis IV engines RT (Oct 6).
What's genuinely new: an agency-level nuclear framework — the sibling of the DoD demo wave (Army leases, Aberdeen IMSR, the Antares space reactor in the newgen thread) and the first NASA-DOE instrument since the fission-surface-power procurement era. No MOU text, scope, or instrument list in our feed — read the primary before weighing.
Watch: the MOU's text (ORNL/NNSA in signatories?), a fission-surface-power RFP, budget follow-through, wire pickup of the "Nuclear NASA" frame.
Links: Isaacman on X
5. Amazon Leo's 1,000th satellite — production solved, rockets are the constraint
What happened: Amazon confirmed its 1,000th Leo satellite built — production running ~27/week, ~400 in orbit — with service rollout "weeks away" and the 2026 service-entry target reaffirmed in coverage. The frame the coverage converges on: satellites outpace launch availability ("Amazon a un problème… c'est d'avoir assez de fusées").
What's genuinely new: the production number is the constellation race's cleanest comparable (vs Starlink's weekly cadence and the V3 ramp), and the bottleneck's public statement is new — exactly what the thread's Ariane-6-consolidation math predicted, with the F9 rideshare exit (post-2028 per the ARK frame) tightening the market Leo has to buy from.
Watch: rollout date (ledgered, review-by Oct 26), a third-provider buy, per-satellite V3 comparisons.
Links: milestone (feed cluster, multi-outlet)
6. Orbital compute gets its balance-sheet layer: $40B in debt for NVIDIA GPUs
What happened: SpaceX is reportedly seeking $40B in new debt to fund NVIDIA chip purchases (the stock slipped on the report), and Morgan Stanley's note now treats the "AI compute business" as a valuation driver.
What's genuinely new: the orbital-data-center thesis arriving as financial plumbing rather than a demo: GPU debt at a scale comparable to the spaceport's capital plan, with the "AI satellites" Gen3 spec (250 kW/satellite, trade-grade unverified) as the hardware side. SpaceX's compute flank now has money, spec, and launch cadence on the table — none of it independently verified.
Watch: the debt instrument's terms, what Gen3's onboard compute actually serves on orbit, Morgan Stanley's AI-compute assumptions.
Links: report (feed cluster; CNBC-grade pickup, primary unverified)
7. Crew-12 home after 237 days
What happened: Dragon Freedom undocked Oct 7, deorbited, and splashed down off the Los Angeles coast Oct 8 (NSF live).
What's genuinely new: none — routine rotation executed on schedule, and that's the point: the Dragon retirement-planning record (Starliner-2-2028, the $359M Boeing bridge) continues against a fleet that demonstrably cycles crews in and out.
Links: NSF overview
Brief
- NASA's bulk launch-buy for the $30B Moon base (Bloomberg chain; Rocket Lab/Blue items already covered): 4–5 launcher families (Falcon/Vulcan/New Glenn named), up to 60 tons, "amid reduced SpaceX availability" — instrument shape (NLS-class buy vs task orders) is the open question; SpaceNews ran the requirements panel (mobility/power/surface logistics).
- Blue Origin's Moon-power entry: a 65-ft solar tower for future lunar bases (stale-bucketed; single source — verify against a Blue primary) — the surface-power shelf gains a Blue entry.
- RKLB's analyst cadence: Synspective's 20-launch deal through 2031 (order book past 100 missions), Citi and Barclays initiations (Buy/Overweight), ARK adding on the Starship-capacity-squeeze thesis, Morgan Stanley "undervalued launch play" — the market layer's verdicts stacking on the Falcon-9-endgame thesis; rideshare exit "after 2028" keeps Neutron's clock central.
- Korea's Nuri fifth launch: record 15 satellites, five LG affiliate components (Innotek camera modules, battery packs) flying as a joint demo — the national-launch + supplier-in-space pairing.
- Vandenberg cadence: SDA's fourth Tranche 1 launch finally lifting after two scrubs; sonic-boom advisories for SoCal.
Dropped
- The SPCX/RKLB daily stock-churn cluster (~100 articles): "should you buy," price-target roundups, Cramer takes, trillionaire-watch, the "$1,000 in 2030" think piece, the Russia/Asia syndications of the same items — the market layer's substance is recorded (Musk's <5% government line ledgered; the $40B report above; MS's note); the daily noise is below the digest bar. The lockup-expiry date (Oct 9) is the only checkable item and passed in-window with no event beyond price action.
- Ship-40 return social-footage pile (~15 items) — the arrival is one event; recorded once.
- "SpaceX used tourist videos as evidence after Starship exploded" — stale-bucketed, single low-tier item, flight/investigation context unverifiable in-window; skip.
- World Space Week listicles / launch-preview reposts / ARM & HAMMER sponsor content — evergreen/promo.