HPE sets 2026 lunar AI and HPC tests on Astrolab rover to Moon's south pole

Hewlett Packard Enterprise intends to execute the first high-performance computing benchmarks on the lunar surface, using a specialised fanless computer module aboard a commercial rover. This initiative aims to process data directly on the Moon rather than transmitting it to Earth for analysis, supporting future lunar exploration efforts.

The hardware, known as Spaceborne Computer-4, will travel to the lunar south pole in late 2026. It is part of NASA’s Commercial Lunar Payload Services programme and will be delivered by a Voyager Technologies Griffin-1 lander. The computer will operate on Astrolab’s FLIP rover, enabling real-time testing of computing capabilities in an extraterrestrial environment.

What we know

HPE states that no other organisation has conducted such benchmarks on the Moon, positioning this as a unique milestone. The system will run High Performance LINPACK and High Performance Conjugate Gradients tests, alongside artificial intelligence evaluations. These AI tests include large language model retrieval-augmented generation, vision inference, and LiDAR processing for mapping and navigation.

Scott Wiest, HPE’s global sales chief technologist, emphasised that future lunar exploration depends on turning data into insights at the point of discovery. The company has developed a software suite to manage the connection to Earth-based systems and ensure the secure delivery of results. This approach brings advanced computing to the “lunar edge,” reducing reliance on ground stations for immediate data processing.

The computer module is lightweight, self-contained, and designed specifically for space environments. Unlike units used on the International Space Station, this device is built to withstand lunar conditions. It measures approximately 10 inches by 4 inches, similar to a standard graphics card, and comprises two component cards. The chassis is engineered to dissipate heat into the rover’s plate, eliminating the need for fans.

Inside the module is 1TB of Kioxia storage, a low-power CPU, and a small GPU. HPE senior director Norm Follett previously described the hardware on the company’s Technology Now podcast. While specific performance targets, power budgets, and chip vendors remain undisclosed, the benchmark results will eventually provide clarity on the system’s capabilities in a low-gravity, high-radiation setting.

Defender's checklist

• Verify the integrity of software updates before deploying them to remote or isolated computing systems.
• Monitor communication links between edge devices and ground stations for unusual data patterns or latency spikes.
• Ensure that AI models running on edge devices have strict access controls to prevent unauthorised inference requests.
• Regularly audit the configuration of data transmission protocols to maintain secure delivery of results to Earth.

The background in plain language

Edge computing involves processing data near where it is generated, rather than sending it to a centralised cloud or data centre. This reduces latency and bandwidth usage, which is critical in environments with limited connectivity, such as space missions. High-performance computing benchmarks measure how quickly a system can solve complex mathematical problems, providing a standardised way to compare computing power across different platforms and hardware configurations.

Background: Large language models

Large language models are zero-shot reasoners only in appearance. They predict the next word by matching patterns in training data, not by applying logical rules. This means they can solve novel problems if similar examples exist in their data, but they fail when true novel logic is required.

Read the full guide: Are Large Language Models Zero-Shot Reasoners? The Reality Check

What to do and how to stay safe: HPE

• Review your organisation’s edge computing strategy to identify potential vulnerabilities in remote data processing.
• Implement robust encryption for data in transit between edge devices and centralised systems.
• Establish clear protocols for updating software on isolated or hard-to-reach hardware components.
• Conduct regular security assessments of AI models deployed on edge devices to detect biases or anomalies.

Step-by-step guide: AI Governance for Small Teams: Practical Controls and Cost

General security guidance from the TechWire Beat newsroom. It is not confirmed advice from the organisations named in this story.

Frequently asked questions

When will HPE’s lunar supercomputer mission launch?

The launch is scheduled for late 2026 as part of NASA’s Commercial Lunar Payload Services initiative.

What benchmarks will be run on the Moon?

The system will run High Performance LINPACK (HPL), High Performance Conjugate Gradients (HPCG), and various AI tests including LLM and LiDAR processing.

Is this the first time supercomputer benchmarks have been run on the Moon?

Yes, HPE states that no other company or organisation has conducted such benchmarks on the lunar surface.