The space shuttle used a hardened 386 (or something similar) that cost a bazillion dollars. The drone currently on Mars used commidity off the shelf chips (snapdragon IIRC). The harsh conditions thing is not something to laugh at, but I also think we might have over-engineered somethings a bit.

> The space shuttle used a hardened 386 (or something similar) that cost a bazillion dollars.

Their main computers were a custom variant of the IBM 360 mainframe. They also duplicated it 5x: 4x with majority vote for error correction, plus a backup running separately-developed software.

https://en.wikipedia.org/wiki/IBM_System/4_Pi

> The drone currently on Mars used commidity off the shelf chips (snapdragon IIRC). The harsh conditions thing is not something to laugh at, but I also think we might have over-engineered somethings a bit.

It should be noted that drone is 1) an engineering experiment and 2) no one will die if it malfunctions, so they can afford to be a bit sloppier.

SpaceX is also using commodity hardware, but they duplicate it 3x with majority vote for error correction.

do you happen to have pointers evidencing this handy? IS this also true for Dragon?

It is, see https://stackoverflow.blog/2021/12/27/dont-push-that-button-...

Thanks! I still hope to learn how they deal with infamous "bi directional requirements Tracing" and similar bureaucracies in the name of dependability. How do they evidence engineering scrutiny, how do the embed it into their processes in such a way that they _evidence_ scrutiny and deliberate engineering _without_ massive paper safety tigers, _without_ tons of slow bureaucracy?

Some of it is classified but you can often find info on certain open sites at NASA or SNL.

The surface of Mars being further from the sun and inside an atmosphere has vastly fewer issues with radiation.

In both cases you're well protected against the solar wind by distance and atmosphere in the case of the Mars rover and by the Van Allen belts in the case of the Space Shuttle. But both still have to worry about cosmic rays. Mars's thin atmosphere helps a bit with that but not nearly as much as you'd like.

EDIT: I suspect a lot of what lets you use Snapdragons in the rover is the idea of making error resistant systems rather than error resistent devices. The same thing that let us replace a single $10,000 hard driver with ten $100 hard drives in a RAID with the same performance and reliability despite the individual drives being much worse.

That’s not really accurate the sun causes real issues in LEO. See chapter 2 starting on page 9 here:

https://commons.erau.edu/cgi/viewcontent.cgi?article=1101&co...

As an anecdotal bit of support, I seem to recall recently SpacX loosing a number of satellites in their LEO constellation because of sun activity

Yes, but that was not due to EM activity, it was because the solar activity made the atmosphere go up higher than usual which meant that the atmospheric drag was too strong for the engines to overcome.

Biggest difference is the lack of risk to human life, tbh

But more issues with temperature swings—ideally radiation wouldn't be the main concern here in Earth's atmosphere.

Rovers regulate internal temperatures to keep electronics happy. Ex: https://mars.nasa.gov/mer/mission/rover/temperature/

We can't teleport something from one atmosphere to another, though. It's a long trip.

That’s a risk, but the chip is off for the trip which makes a significant difference. Fail safe is much easier if you can run diagnostics without needing the system to be doing it’s job.

Also, human lives aren’t at stake if things go awry.

I designed a drop in replacement ignition ECU for Ferrari 308 cars from the 1980's (the Magnum PI model), as after almost 40 years a lot of them are now failing. When I started the project, I had no idea how miserable an environment the car actually was. It wasn't hard to get a prototype that basically worked, but I had to go through quite a few revisions before I got something that could withstand the torture of everyday use (and things like electrically dirty jump starters that would completely fry my units until I put a protective power supply in it against +/-100VDC surges.)

My daily is also an 88. It's running very standard Bosch of that era. At some point I took my 93 (same model) and swapped it with newer fuel management, ecu, etc out of a 2005 Saab. Same valve cover allows you to the stock coils and setup, a timing sensor solves that, and boom, good to go.

Well sorta, turns out the 88 is far more reliable and consistent - except for a few hard to find parts... namely the distributor is a total piece of crap after that many years... and it's not very efficient. I get 50% more performance and 50% more gas milage on the uprated one.

Difference in reliability is probably... once every 25k versus once every 50k though.

Those bosch systems were all over volvo, saab, porsche, VW, Audi, etc and are dead reliable but not very hackable.