Space Industry Cheat Sheet: Golden Dome Gets $10B Bigger While Artemis Heads Back to the Pad
The space sector doesn’t take breaks, and this week delivered a full slate of stories that matter — from a missile defense budget that just got bigger to a moon rocket that finally made it back to the launchpad. Buried in the news cycle were two institutional accountability moments the defense acquisition community should not overlook. Here’s what happened, why it matters, and what to watch.
1. Golden Dome Gets a $10 Billion Budget Upgrade — Because Space Is Now the Foundation
The headline number is hard to ignore: the Pentagon has increased the cost estimate for the Golden Dome missile defense architecture by $10 billion, bringing the total program estimate to $185 billion. The driver isn’t bureaucratic bloat — it’s the accelerated demand for space-based sensing, tracking, and data transport capabilities that the original estimate underweighted.
Three specific areas are driving the increase:
- HBTSS (Hypersonic and Ballistic Tracking Space Sensor): The on-orbit system designed to track hypersonic glide vehicles from their midcourse phase through terminal approach
- Space Force Data Transport Layer: The communications backbone that connects sensors to shooters in near-real time
- Space-Based AMTI: Moving target indicator capability applied to airborne threats, not just ballistic missiles
What’s most interesting this week is the formalization of the Space Data Network (SDN) — the integrated communications architecture that makes all of this actually work together. The SDN pulls together proliferated LEO and MEO SATCOM constellations into a coherent command and control pathway, turning satellite data into actionable targeting information fast enough to matter. This isn’t a concept paper anymore. It’s a program.
Alongside the budget news, Space Systems Command awarded a $446.8 million Ground Management and Integration (GMI) agreement to Kratos Technology & Training Solutions on March 17 to support the launch and operations of Epoch 1 and Epoch 2 for the Resilient Missile Warning and Tracking (MWT) architecture in Medium Earth Orbit. The first Epoch 1 satellites are expected to launch in the second half of 2026.
Industry also demonstrated the command and control (C2) layer for Golden Dome this week — a live consortium demonstration from nine defense firms confirming the architecture is on track toward initial operational capability in 2028, with full objective capability targeted for 2035.
Why it matters: This isn’t just a missile defense program anymore. Golden Dome is becoming the organizing framework for U.S. space-based ISR, tracking, and battle management. The $185 billion number will get headlines, but the operational architecture underneath it is what actually changes the deterrence calculus. When the SDN is operational and the MWT constellation is in orbit, we’ll have persistent, all-domain threat tracking from space at a scale that currently doesn’t exist. That is worth every dollar — if the program doesn’t become a “boondoggle of bad ideas” in execution. The Kratos award and the C2 demonstration suggest the program is moving with actual urgency. Watch the Epoch 1 launch timeline closely.
Sources: Air & Space Forces Magazine, Defense Scoop, Space Systems Command, Air & Space Forces Magazine — C2 Demo
2. Artemis II Is Back on the Pad — And April 1 Is a Real Launch Date
On the night of March 19 into March 20, NASA’s Artemis II Space Launch System and Orion spacecraft made their second rollout to Launch Complex 39B at Kennedy Space Center. The 4‑mile journey from the Vehicle Assembly Building started around 12:20 a.m. EDT and took roughly 11 hours.
This is the second time the vehicle has made this trip. The original rollout on January 17 was followed by a return to the VAB on February 25 after engineers identified a helium flow issue in the rocket’s upper stage during a wet dress rehearsal. Additional maintenance and system retesting were performed before the team was confident enough to return to the pad.
With the rocket now at the launchpad, NASA is targeting a wet dress rehearsal ahead of a launch window that opens April 1, 2026, and extends through April 6. The crew — Commander Reid Wiseman, Pilot Victor Glover, Mission Specialist Christina Koch, and Canadian Space Agency Mission Specialist Jeremy Hansen — entered quarantine in Houston on March 18.
Artemis II will be the first crewed flight of the Artemis program: a 10-day journey around the Moon and back. No lunar landing — this is the deep-space shakeout of the human-rated SLS/Orion stack before anyone steps on the surface.
Why it matters: The original Artemis schedule has slipped so many times that the program’s credibility has taken real hits. Getting back to the pad after a real technical problem was resolved — not waived — is the right answer. The April window is legitimate; the crew is in quarantine. If the wet dress rehearsal clears without major findings, we’re looking at humanity’s first crewed deep-space flight since Apollo 17 in December 1972. Fifty-three years is long enough. Watch for the wet dress rehearsal results this week.
Sources: NASA, Space.com, Spaceflight Now
3. NVIDIA Brings Data-Center-Class AI to Orbit
At GTC 2026, NVIDIA CEO Jensen Huang announced the company’s entry into space computing with a new suite of hardware purpose-built for the orbital environment. The headline product is the Space‑1 Vera Rubin Module, a computing platform designed to deliver up to 25 times more AI compute than NVIDIA’s H100 GPU while operating within the size, weight, and power constraints of a spacecraft.
NVIDIA is also fielding the IGX Thor and Jetson Orin platforms now — the Space‑1 module is still coming. Current customers already using NVIDIA’s space computing technology include Aetherflux, Axiom Space, Kepler Communications, Planet Labs, Sophia Space, and Starcloud.
The problem these platforms solve is real: satellites generate enormous volumes of data daily, and today all of it has to be downlinked to Earth for processing. That creates latency, bandwidth bottlenecks, and dependence on ground infrastructure. Moving AI inference on-orbit — processing imagery, signals, and sensor data before transmission — cuts latency, reduces downlink requirements, and enables faster autonomous decision-making in space.
Why it matters: For the commercial space sector, this is a transition point. In-space computing has been a near-term aspiration for years; NVIDIA’s hardware commitment with named customers signals that the economics are getting real. For the defense side, on-orbit AI processing is directly relevant to ISR, missile tracking, and space domain awareness — the same mission threads running through Golden Dome. Expect defense contractors to be the next wave of customers once reliability data accumulates. Pay attention to what Planet Labs does with this; they are the canary in the coal mine for applied orbital AI.
Sources: Payload Space, NVIDIA
4. The Pentagon’s Top Weapons Tester Has Notes on ATLAS
The Director of Operational Test & Evaluation (DOT&E) released its annual report on March 16, and the Space Force’s Advanced Tracking and Launch Analysis System (ATLAS) earned an uncomfortable chapter.
ATLAS was approved for initial operations by the Space Force last September, described at the time as a “revolutionary leap forward for our warfighters.” DOT&E’s report, covering testing conducted prior to that acceptance, found that ATLAS “as tested, did not contain the minimum viable capability necessary for SPADOC decommissioning” — where SPADOC is the Space Defense Operations Center, a 1980s-era command and control system the Space Force has been trying to replace for years.
L3Harris is developing the program, with Space Systems Command serving as the lead integrator. ATLAS marks the initial phase of a larger Space C2 modernization initiative. A Space Force official told Air & Space Forces Magazine that Combat Forces Command has a “deliberate plan” to decommission SPADOC this year but did not provide details about the remaining deficiencies.
Why it matters: Software integration issues have cost this program nearly three years and already placed it on the then-acquisition leadership’s watchlist. The DOT&E report isn’t claiming that ATLAS fails — it’s saying it wasn’t ready to replace SPADOC when the service approved it. The Space Force’s credibility hinges on whether the deliberate decommissioning plan includes enforceable steps or is just a timeline that is likely to slip again. C2 modernization is fundamental to everything else the Space Force aims to achieve, including Golden Dome integration. You can develop all the sensors you want, but if the C2 layer doesn’t perform reliably, the operational picture collapses.
Source: Air & Space Forces Magazine
5. Two Quiet Wins the Space Force Doesn’t Advertise Enough
Two other stories from the week deserve more attention than they got.
EPS‑R Achieves Operational Acceptance: On March 20, the Space Force announced that the Enhanced Polar System – Recapitalization (EPS‑R) achieved operational acceptance. EPS‑R provides secure, jam-resistant satellite communications coverage in the Arctic — the same region where contested peer competition is increasingly focused. This extends protected polar SATCOM capability into the 2030s. It isn’t flashy, but the polar region SATCOM gap has been a real vulnerability. Closing it matters.
GPS III SV-10 Moving to SpaceX Falcon 9: The final GPS III satellite will now launch on a SpaceX Falcon 9 after United Launch Alliance’s Vulcan rocket was grounded in February due to persistent solid rocket booster issues. GPS is foundational infrastructure — for the military and everyone else. Moving the launch to a proven vehicle was the right call. ULA’s Vulcan reliability issues remain a real program management story worth tracking.
Sources: Space Force, Breaking Defense
What to Watch Next Week
- Artemis II wet dress rehearsal results — this is the final significant technical gate before the April 1 launch.
- Golden Dome Congressional response — $185 billion invites appropriators to ask hard questions; watch for hearings or statements
- ATLAS decommissioning timeline — will the Space Force publish a specific schedule or keep it vague?
- ULA Vulcan investigation updates — two GPS launches now redirected; the question is whether Vulcan returns to flight in 2026 at all
The thread running through all of this is integration, not as a buzzword, but as a hard operational requirement. Golden Dome only works if the SDN moves data fast enough. ATLAS only matters if it actually replaces SPADOC. Artemis only validates the architecture if the wet dress rehearsal does not surface another technical ghost. NVIDIA’s on-orbit compute only changes the ISR calculus if the defense community moves fast enough to adopt it. EPS‑R and GPS III SV-10 are already doing what they are supposed to do, and nobody is writing headlines about them, which is exactly the point. This business rewards programs that perform quietly and punishes those that generate press releases before they earn them. The space sector is building the architecture that will define the next fifty years of American deterrence. What happens over the next few weeks, on the launchpad, in front of appropriators, and in the C2 software stack, will tell us whether we are ahead of that timeline or behind it. We cannot afford to be behind it.
Pax ab Space
Clinton Austin is a Senior Business Development Director for GDIT who covers the U.S. Air Force, the U.S. Space Force, and the Missile Defense Agency.
The views expressed are those of the author and do not necessarily reflect the official policy or position of General Dynamics Information Technology.
March 23, 2026