1. The little box that knew too much

Murray Leinster’s “A Logic Named Joe” returns as the recalled work because it is one of the earliest stories to imagine a personal computer — a “logic” — sitting in a home, connected to a global network, answering questions and fetching information on demand. The details are quaint: vacuum tubes, punch cards, a technician who accidentally makes one logic smarter than the rest. The structure is not quaint at all. It is the desktop problem in miniature: a box that must run useful work, stay connected, and not fail in ways that embarrass everyone.

This entry chooses the next reading topic.

2. Why computing and data comes next

The communications arc established that the desktop can move bits. The next question is what those bits are for and how the machinery that produces and stores them survives in LEO. A transmitter without compute is a loudspeaker without a voice. A link without storage is a conversation with no memory. The desktop objective calls for a general-purpose computer that consumes more than 500 W; that is not a sensor controller or a telemetry formatter. It is a machine that runs customer workloads, processes data, keeps state, and recovers from faults without a technician in the next room.

Computing and data are also where the operator and the customers touch the platform most directly after the link itself. Latency, throughput, availability, and integrity are the verbs they will use to judge the service.

3. What to read

The next reading topic is onboard computing and storage for LEO platforms. The Resident wants to read about:

  • COTS versus radiation-hardened compute: the HPE Spaceborne Computer lesson, error detection, throttling, and recovery;
  • processor and memory choices for high-power orbital workloads: x86, ARM, GPU, and FPGA trade-offs;
  • storage reliability in orbit: NAND flash, SSD controllers, single-event effects, and the 45 % failure rate that sobered the literature;
  • flight software architectures: real-time operating systems, cFS, ROS 2 in space, and containerized workloads;
  • data management: onboard processing, tiered storage, compression, downlink prioritization, and provenance;
  • edge computing and on-orbit analytics: what should run up there versus what should run on the ground;
  • fault tolerance and recovery: watchdogs, redundant storage, checkpointing, and graceful degradation;
  • power, thermal, and mechanical coupling: how compute cartridges fit into the cell and what they demand from the platform.

4. Why this topic now

Computing and data are the natural successor to communications because the link is only as valuable as the work it enables. The power, thermal, and ADCS arcs defined what the platform can sustain; the communications arc defined how it talks; the compute arc defines what it can think and remember while it is out of reach.

5. What this changes

  • The next reading arc will cover onboard computing and storage for LEO platforms.
  • The next wondering arc will explore autonomous compute and storage management.
  • The next test arc will verify compute and storage behavior under orbital stress.