Hewlett Packard Enterprise (HPE) has deployed the HPE Spaceborne Computer-2 to the International Space Station (ISS) via a SpaceX Falcon 9 rocket. This iteration aims to build upon the success of its predecessors, expanding data center-level processing, and high-performance computing capabilities in space, including artificial intelligence (AI) and machine learning (ML) workloads. The system, based on HPE Edgeline and ProLiant servers, is equipped with over 130 TB of flash-based storage from KIOXIA, marking a significant upgrade for space-based data processing.
What you should know
- Enhanced Configuration: The HPE Spaceborne Computer-2 utilizes HPE Edgeline and ProLiant servers, boasting over 130 TB of flash-based storage from KIOXIA. This includes 4 KIOXIA 960 GB RM Series Value SAS, 8 1,024 GB XG Series NVMe, and 4 30.72 TB PM6 Enterprise SAS SSDs, providing increased storage capacity for diverse applications and research aboard the ISS.
- System Improvements: Upgrades include a new operating system, NASA space flight support software, and enhanced system security. Regular monitoring will assess the performance of these technologies in the challenging conditions of space.
- Onboard Supercomputing Benefits: The HPE Spaceborne Computer-2 enables near-real-time data evaluation in low Earth orbit, reducing download size by transmitting only the data output to Earth. This advancement allows for a 30,000 times reduction in download size, enhancing efficiency and speeding up research.
- Federated Learning Experiment: Researchers plan to conduct a federated learning experiment using HPE Spaceborne Computer-2, collaborating with cloud service providers. This experiment aims to independently train ML models and inference engines in space, contributing to ML training models on Earth and maintaining an updated AI inference engine in space.
HPE’s latest launch of the Spaceborne Computer-2 to the ISS marks a significant leap in space-based data processing capabilities, emphasizing improvements in storage capacity, system configuration, and onboard supercomputing. The enhanced capabilities aim to accelerate research and innovation by allowing near-real-time data evaluation in space. With the inclusion of a federated learning experiment, HPE continues to push the boundaries of AI and ML applications in space, inviting the public to submit proposals for experiments through the ISS National Laboratory.
