Apollo 11 vs. USB-C Chargers (2020)

Spaceflight computing: past, present, future

• Debate over whether legacy CPUs like the Z80 are still/will be used; some recall historical use, others working in current space projects say modern ARM, PowerPC (RAD750), and emerging RISC‑V parts dominate.
• One comment cites NASA moving to new RISC‑V–based spaceflight processors; another notes RAD750 is common but not universal due to cost/power, with many missions qualifying automotive/industrial‑grade hardware instead.
• Discussion of data links: older protocols like Kermit/Xmodem were used locally (e.g., ISS), but deep‑space and lunar comms largely use CCSDS standards.
• SpaceX is cited as using commodity x86 with multiple cores in lockstep and triplicated “strings” that vote, offloading actuation to microcontrollers.

Apollo Guidance Computer and Apollo-era systems

• AGC considered state‑of‑the‑art for its size, power, and reliability, though obsolete by late 1960s compared with newer aerospace machines.
• Its 2 ft³ volume constraint was negotiated before the relevant ICs even existed.
• Comparison to Data General Nova and IBM 4Pi highlights AGC’s strong performance/weight trade-off.
• Apollo used additional computers: analog Flight Control Computer (FCC) in Saturn V, plus human “computers” on the ground.
• The Lunar Module guidance programs had several modes; an automatic landing mode existed but was never actually used in flights.

USB-C chargers and ubiquitous embedded CPUs

• Many USB controllers and even simple peripherals embed fully programmable CPUs; microcontrollers are now cheaper than custom ASICs for many roles.
• This leads to reflections on “software eating the world” and whether the role of pure hardware/EEs is shrinking; others argue EEs remain essential wherever systems meet the physical world.
• Some security concern raised about “wall warts” and malicious smart cables.

Analog vs digital computing and Turing completeness

• Extended debate about whether analog flight control hardware should be called “computers,” and what distinguishes a “computer” from mere arithmetic hardware.
• Several commenters discuss analog, digital, and hybrid architectures, and whether general‑purpose analog computers can be Turing‑complete; references to theoretical models (e.g., GPAC) and limits due to precision and physics.
• Consensus: analog systems compute, but mapping that to modern “general‑purpose computer” terminology is contentious.

Radiation, reliability, and redundancy

• Discussion of rad‑hard vs shielding vs redundancy.
• Some argue shielding plus COTS parts (with reset/reboot strategies) is viable for many missions; others emphasize dedicated rad‑hard parts are disappearing from major vendors.
• Multi‑way voting (SpaceX rockets, fly‑by‑wire aircraft, Saturn LVDC triple‑modular redundancy) is dissected; there is disagreement on how to count “N‑way redundancy” and how to avoid single points of failure in the voter.
• Parallel, independently developed hardware/software stacks are noted as a way to mitigate common‑mode failures, though one cited study suggests independent teams may still create correlated bugs.

Compute progress, everyday tech, and skepticism about comparisons

• Many find it striking that charger/controller chips rival or beat Apollo‑era computers; others warn against over‑interpreting this as making lunar missions “easy” today.
• Comments highlight that Apollo’s success depended on exceptional engineering, ground support, and bespoke systems, not just CPU speed.
• Some extrapolate Moore’s‑law‑style curves to imagine future cables with cluster‑level AI capability; others joke that software bloat will consume gains.
• Debate over whether governments still enjoy a distinct computing edge; some think big tech firms now match or exceed most nation‑state compute except at specialized supercomputing centers.