Re: [emacs-tangents] What may be the design constraints on computing going to Mars?

[Van Ly]

Jean Louis <bugs@static.rcdrun.com> writes:

                       Assuming the compute workload required over a 24
>> month journey span is 1x serial and 9x parallel, a 4x cluster of
>> 
>>     - 4x Nvidia DGX Spark for parallel
>>     - 1x something else in the form factor for serial
>> 
>> radiation hardened is enough for the first attempt.  WDYT?
>
> I don't. I think is waste.  I think all that money could be used to give to 
> people who are really lacking food and who are under the depopulation agenda 
> strategy.
>

That money isn't a lot, knowing there are governments, such as, Uganda, France, 
Russia and China, that are more inefficient than USA.  The DOGE says, in the 
USA, a political actor will start an NGO with $100, and all of a sudden, $2 
billion will be stripped from the one government fund through fraud and waste 
to keep feeding it.  Or, what must be a $10 thousand survey on government 
recreational park services, is actually billed at $2 billion.  And, no further 
processing or followup action is done on the information collected in the 
survey.

In fewer than 400 words, what does your preferred LLM model say, about the 
depopulation agenda starvation strategy, you see and want to prevent, as having 
higher priority than Starship going to Mars in 12 months?  I haven't come 
across this idea ,at all.

-- 
vl