Hacker News Reader

Comment by hga | Hacker News Reader

Google helped me find this very detailed article on the probe: http://www.spaceflight101.com/hayabusa-2.html

Which says the solar panels are designed to produce 1,400 watts at the furthest distance from the sun. So it gets this 200 Mhz 64 bit MIPS chip, whereas the ~200 watts at Pluto New Horizons uses a 12 Mhz 32 bit MIPS chip (and the power gets steadily lower as it heads out and hopefully looks at a Kuiper belt object).

View on HN

I don't know anything about architecture or power usage of these CPUs, but it wouldn't even surprise me if that 200MHz CPU took less power than that 12 MHz one.

There also is a window of almost 9 years between the launch dates of these systems (January 2006 vs December 2014)

That's six times the 18 months of Moore's law, or a factor of 64, so it could explain much of the difference.

Maximum TDP for HR5000 is 4-6W (fact) whereas the Mongoose-V might be in the sub 1W range (speculation).

I have a feeling Mongoose-V (32-bit MIPS) is manufactured in a 0.18um process node. I'm not 100% sure but it would coincide with the time of the release. HR5000 (64-bit MIPS) might be manufactured in a more recent SOI process node.

I don't think these are normal CPUs. I'm not an expert on this by any means, but is a radiation-hardened CPU (as is needed in space) manufactured with a different process than chips designed to operate within the Earth's atmosphere?

Yes but process node characteristics such as the gate length and/or the silicon film thickness should be relatively similar.

Reply on news.ycombinator.com