Full Breakdown
Hewlett Packard Enterprise's (HPE)’s Supercomputer Set to Operate on the Moon’s South Pole via Astrolab’s FLIP Rover
By Drooid · · How we work
Core Mission Overview
Astrolab’s Flex Lunar Innovation Platform (FLIP) rover will carry Hewlett Packard Enterprise’s Spaceborne Computer-4 (SBC4) to the lunar south pole under NASA’s Commercial Lunar Payload Services (CLPS) program. The rover will be delivered by Astrobotic’s Griffin-1 lander on a SpaceX Falcon Heavy launch later this year, marking the first deployment of a high-performance computing (HPC) and artificial-intelligence (AI) system on the Moon’s surface.
Technical Design and Objectives
SBC4 is a compact, fan-less module built from heat-conductive aluminum. It houses a low-power 64-bit CPU, a GPU, and a pair of mirrored Kioxia solid-state drives (~1 TB total). Power comes from onboard batteries supplemented by solar panels, and the rover’s cooling system manages thermal loads.
Primary objectives:
- Verify survival through extreme temperature swings, including up to 150 hours of continuous darkness.
- Run an AI services container demonstrating scalable AI applications in situ.
- Execute large language-model retrieval-augmented generation (RAG) tests and machine-learning algorithms on camera and LiDAR data for autonomous navigation and terrain mapping.
- Perform HPC benchmarks such as High-Performance Linpack (HPL) and High-Performance Conjugate Gradient (HPCG), potentially yielding the first Top500 submission from a celestial body.
A Jet Propulsion Laboratory (JPL) localization software experiment, previously used on Mars, will match onboard terrain images against a stored map, enabling GPS-free navigation.
Data & Statistics
- SBC4 dimensions: ~5 × 10 inches (? 12.7 × 25.4 cm), ~1 inch thick.
- Storage: two Kioxia BG6 M.2 SSDs in RAID 1, ~1 TB.
- FLIP rover size: ~7 ft (? 2.13 m) long, ~½ ton.
- Wheels: large, air-less pneumatic-style wheels developed by Venturi Space.
Official Statements & Responses
Astrolab CEO Jaret Matthews highlighted the challenge of designing a platform that must operate autonomously for a full lunar year, survive multiple dark periods, and travel hundreds of kilometers. He said the mission will show how commercial lunar mobility can support technology development and scientific research.
Chief scientist Mark Fernandez noted that SBC4 relies largely on off-the-shelf components protected by software-based redundancy, with continuous hardware monitoring and automatic throttling to mitigate radiation-induced upsets.
NASA’s CLPS program frames the flight as a testbed for emerging technologies that could later be integrated into crewed Artemis missions, especially for edge-computing tasks that reduce the need to transmit raw data back to Earth.
Why It Matters
Demonstrating AI and HPC on the Moon could enable future crewed missions to rely on autonomous decision-making, reducing latency-induced delays and bandwidth constraints that limit real-time control. Successful operation of SBC4 would provide a proof of concept for lunar edge-computing platforms, informing the design of larger systems for sustained lunar habitation and commercial activities.
Conflicting Reports & Gaps
Specific details about the CPU, GPU, and AI model sizes remain undisclosed, as do the exact scientific payloads and the schedule for post-landing data releases.
What’s Next
The FLIP rover and SBC4 are slated for launch later this year on a Falcon Heavy, with a planned year-long operational period at the lunar south pole. Data from the AI and HPC experiments will be analyzed to inform the design of future Artemis-compatible lunar platforms and assess the feasibility of larger edge-computing installations on the Moon.