NASA wants to send a computer to the moon
For short-duration missions to the lunar surface, NASA needs a computer that can process and store data and sequence tasks for multiple payload subsystems.
The system must be able to withstand radiation-induced errors, operate in a vacuum and over an extended temperature range of -40C to +70C while on the lunar surface. Additionally, the Linux-based computer system must withstand all of the stress of launch, transit, decent to and roving on the moon.
NASA's request for information suggests minimum data storage of 128MB and 128MB of memory with error detection and control. The minimum processor speed has not been determined, but NASA-led software development efforts have shown a CoreMark Score of 600 is sufficient. The space agency asks vendors to specify mass, volume and power requirements.
The project has a Class D rating, which means the agency will accept designs with higher risk and lower tolerance.
In August, NASA sent an HPE Linux supercomputer to the International Space Station to give astronauts high-performance computing capabilities for onboard operations and provide extra computing power for their experiments. The HPE supercomputer uses 2-socket “pizza-box” servers from the HPE Apollo 40 family with Broadwell-class processors and a 56 gigabit/sec interconnect.
The year-long experiment will test whether off-the-shelf servers and custom-built software can operate in space without hardening them against radiation damage. HPE plans to protect the computer by slowing it down when there's a solar flare or other radiation hazard, the company explained.
Currently, many of the calculations for ISS research projects are performed on Earth then transmitted to space. While this method works well for transmission in low Earth orbit or to the moon, as astronauts travel farther from home, communications latency will grow, limiting efficiency and possibly endangering crew members if they face a situation they can't resolve themselves.