# SpaceX SDN Backbone Award On **May 26, 2026**, the Space Force awarded [[SpaceX]] **$2.29 billion** to build the **Space Data Network (SDN) Backbone** — a proliferated low-Earth-orbit constellation of optically interconnected satellites that will move military data **through space** rather than routing it through terrestrial relay networks and ground stations. A fully operational prototype is expected by the **end of 2027**. This is the plumbing of American military communications, and one company is building it. ## Terms | Item | Detail | |---|---| | Announced | May 26, 2026 | | Awarding office | [[Space Systems Command]] | | Value | **$2.29B** | | Instrument | **Firm-fixed-price Other Transaction Authority** delivery order | | Scope | Proliferated LEO constellation, optical crosslinks, high-capacity low-latency transport | | Prototype target | End of 2027 | | Former program name | **MILNET** | ## It used to be called MILNET, and Congress fought about it The SDN Backbone is the renamed continuation of **MILNET**, a program first revealed publicly in **2025**: a government-owned, contractor-operated constellation of **480+ satellites** built on [[SpaceX Starshield]] hardware, run jointly by the Space Force and the NRO. Understanding that lineage is what makes the 2026 award legible, because the objections were made on the record before the money moved. The core congressional objection was **over-dependence on a single vendor**. Senator [[Chris Coons]] put it bluntly: MILNET had *"no competition, no open architecture, no leveraging a dynamic space ecosystem."* Members also complained they could not get basic programmatic detail out of the Space Force about a constellation being stood up on their appropriations. The dispute had a concrete fiscal expression. The Pentagon moved to supplant the [[Space Development Agency]]'s Transport Layer — the competitively built, multi-vendor data-transport program — with MILNET, and **halted Tranche 3**. Congress pushed back and **restored roughly $500M to keep Tranche 3 alive**, deliberately preserving an alternative architecture so the department could not, as one analysis put it, press the *"easy button"* of a sole-source giant. That is an unusually clean example of a legislature identifying vendor capture in advance and spending money specifically to prevent it. Whether it worked is now testable: the SDN Backbone award went ahead at $2.29B regardless. ## The 38:1 ratio On **August 13, 2026**, the Space Force awarded **$12M each to five companies — $60M total** — to demonstrate that satellites from multiple vendors could connect into the Space Data Network: **Amazon LEO for Government** (see [[Amazon]]), [[Lockheed Martin]], [[Northrop Grumman]], [[Rocket Lab]], and York Space Systems. Each must field "Space Exchange Point" satellites — effectively routers into the network. Set against the backbone award, the arithmetic is the argument: | Role | Recipient | Value | |---|---|---| | Build the network | SpaceX | **$2.29B** | | Learn to plug into it | 5 competitors | **$12M each ($60M total)** | **A ratio of roughly 38:1 between the incumbent's backbone and the entire multi-vendor interoperability program.** This is what "multi-vendor architecture" means once one vendor already owns the backbone: competitors are not being funded as *alternatives to* SpaceX but as *peripherals of* SpaceX. Diversification language, dependency architecture. Two absences are conspicuous. [[Blue Origin]] and [[United Launch Alliance]] — SpaceX's principal national-security launch rivals and its most likely protest filers — **do not appear on the August 13 award list at all**. ## Significance - **This is where the sole-source thesis actually holds.** Unlike [[SpaceX SB-AMTI Award]], which was competitively awarded against a real vendor pool, the SDN Backbone descends directly from a program Congress objected to *by name* as having no competition. When [[SpaceX Contracts]] argues dependency, this is the strongest available exhibit. - **Transport is stickier than launch.** A launch contract is recompeted every few years and a rival with a working rocket can take it. A **data-transport backbone with proprietary optical crosslinks becomes the network the joint force is built around** — switching costs compound with every system that integrates against it. The Aug 13 awards accelerate that: five competitors are now spending federal money making their satellites compatible with SpaceX's architecture. - **Government-owned, contractor-operated.** The constellation is nominally government property, operated by SpaceX. That structure concentrates operational knowledge — and, in a crisis, operational discretion — in a private firm, which is the same structural exposure documented for [[Starlink]] in Ukraine and for [[SpaceX Starshield]]. - **Attack surface.** Analysts have flagged that a network with hundreds of satellites, many ground stations, and a large user-terminal population presents a broad attack surface; compromise of the software or ground control could disrupt or expose communications across the joint force. - **Awarded pre-IPO.** Like the SB-AMTI award three days later, this landed in the run-up to the **June 12** listing. See [[SpaceX IPO Filing]]. ## Connects to - [[SpaceX SB-AMTI Award]] — the sensing companion to this transport layer - [[SpaceX Starshield]] — the hardware lineage; MILNET was built on Starshield - [[Space Development Agency]] — the competitive Transport Layer this displaced - [[Chris Coons]] — the sharpest on-record objection - [[Golden Dome]] — the architecture both awards feed - [[SpaceX Contracts]] — consolidated ledger - [[Defense Tech Conflicts of Interest]] — the oversight frame - [[SpaceX Watch Log]] — running feed (logged 2026-08-13) ## See Also - [[SpaceX]] — company page - [[SpaceX Dossier]] — dossier front door - [[SpaceX Events]] — index of SpaceX events/deals - [[U.S. Space Force]] · [[Space Systems Command]] - [[Blue Origin]] · [[United Launch Alliance]] - [[Deals Hub]]