Tickline News

Space digest — 2026-10-04

2026-10-04 Space

Top line: Mostly Flight-14 aftermath and analysis, but the window's one genuinely new artifact is SpaceX's Gen3 Starlink spec sheet — 250 kW, 10 Tb/s, onboard edge compute — quantifying the orbital-data-center market's SpaceX flank. Rocket Lab keeps converting its Synspective relationship into flights, and the F9-rideshare wind-down grew an operator-facing memo.

1. Gen3 spec sheet: 250 kW per satellite, 10 Tb/s, onboard edge compute

What happened: SatNews (Oct 4, 8:27 PM ET): technical specifications for SpaceX's next-generation Starlink "Gen3" satellites were disclosed Oct 1 — 250 kW per satellite, 10 Tb/s throughput, onboard edge compute, framed in trade coverage as "AI satellites."

What's genuinely new: the per-satellite numbers behind the Gen3 era are now on the record. They pair directly with Flight 14's Sep 28 deployment of 26 production V3s and matter beyond connectivity: onboard compute is the SpaceX flank of the orbital-data-center market (Google's Suncatcher arithmetic pegged that market at ~1,600 Starship launches; the Gen3 FCC filing's 100k-satellite scale is the mass math). Provenance: trade-grade single source (SatNews); the primary disclosure itself has not passed through our feed — verify against the FCC filing before weighting the 250 kW figure.

Watch: primary-source confirmation; what the onboard compute actually serves (Starlink's own routing/AI functions vs third-party workloads — the difference between a smart router and a data-center node).

Links: SatNews

2. The F9-rideshare wind-down goes operator-facing — and competitors' 2027 answer

What happened: the rideshare freeze (on record since July) re-surfaced via Scanway CEO Jędrzej Kowalewski's X post (Sep 30, ~18k views): SpaceX is ending "retail-level space access" for Transporter/Bandwagon missions — no bookings after 2028, manifest full — "according to information from SpaceX partners cited by SpaceNews and WSJ." No official SpaceX announcement ever came; the trade record corroborates the shape (SatNews Jul 20: SpaceX limits future commercial rideshare bookings past 2028 amid severe manifest saturation; SpaceQ unpacked the smallsat-market/Canada implications). A JITE.news memo (note.com, Oct 4) organizes the claim into operator guidance: secure 2028–2030 launch capacity, note integrators bulk-buying Falcon 9, review contracts within 90 days. Same-day companion (Nextbigfuture): Super Heavy has deployed >50 t to orbit, scaling toward 100–120 t/launch-month — SpaceX's Falcon-9 competitors "might have flight in 2027."

What's genuinely new: the freeze now has an operator-reaction layer — a satellite-components CEO publicly documenting the loss of retail access, and the access-gap arithmetic (who fills 2028+ rideshare demand) appearing in analysis print. The thread's watch — does a Starship-class rideshare product exist before the gap opens — remains unanswered; retail access ends with the 2028 manifest either way.

Links: note.com/JITE.news · Scanway CEO post · SatNews Jul 20 · Nextbigfuture

3. Flight 14 aftermath converges: engine issues, and NASA's 2028 clock

What happened: the analysis layer of Flight 14 (flown Sep 28, orbit + 26 V3s + earlier-than-planned ship return with engine problems — all on record) landed through the window: engine issues "raise stakes for NASA's Moon plan" (digitimes et al.); opinion pieces land the milestone frame ("once you get to orbit, you're halfway to anywhere"); the "fully operational by 2028" piece enumerates what's still untested — refueling, orbital docking, and NASA counting on the timeline. Visual record also landed: the first Starlink V3-deployment picture from the upper stage, live views beamed through Starlink during the flight, and Starbase/Louisiana feature color.

What's genuinely new: nothing hardware-side — this is the analyst layer converging on the record's open list. The load-bearing items are unchanged: F15 NET Oct 19 (catch attempt?), F16 NET Oct 30 (LC-39A debut), and the cadence forecast already cut (Musk: weekly/twice-weekly 2027; Chanos: an 80% downgrade read).

Links: digitimes · fully-operational-by-2028

4. Rocket Lab flies Synspective again — "Owl New World" delivered

What happened: Rocket Lab launched Synspective's Earth-observing radar satellite (mission "Owl New World," Oct 4) — the first execution of the 20-launch Synspective agreement announced Sep 30 (booked total 47, backlog past 100 incl. Kepler's Neutron reservation).

What's genuinely new: execution rather than booking — the record deal is already producing launches. Investor-layer pieces (RKLB technical-analysis riffs on the $200M/200-day line) ride in the same cluster and are dropped below.

Links: MSNBC video/MS syndication

5. SMILE begins science: ESA–CAS publish first magnetosphere images

What happened: after a May launch and a multi-month commissioning period, ESA and the Chinese Academy of Sciences' SMILE (Solar wind Magnetosphere Ionosphere Link Explorer) has officially begun science operations, releasing first UV and soft-X-ray observations of Earth (NASASpaceflight, Oct 4, ~6:50 PM ET).

What's genuinely new: an ESA–China mission passing commissioning cleanly and publishing early — noteworthy as functioning international science cooperation in a beat otherwise dominated by launch-market friction; no recurring thread (single item).

Links: NASASpaceflight

Brief

Dropped