What are NASA's odds for landing on the Moon in 5 years?

[Guest]

First chips, yes, but this wasn't "pushing the boundaries" like Apollo was. ICBMs were deployed in hundreds, the chip technology was new, but unlike Apollo tech, it was ready for widespread deployment. The many uses of computers were well known even then, and physical limits were far away, so rapid growth was to be expected. A jump like Apollo would be like trying to make CPUs matching today's performance in the 80s-90s. They knew it was possible, and could have tried if they wanted, but there was no point, so it wasn't done. Hardware development is driven by improvements in software, which must consider the hardware on which it will run, leading to a rather incremental development model. Moore's law is a consequence of that fact.

Boeing 2707 was killed by a variety of factors, but it didn't have anything to do with NASA. 

Edited May 22, 2019 by Guest

[wumpus]

11 hours ago, Dragon01 said:

First chips, yes, but this wasn't "pushing the boundaries" like Apollo was. ICBMs were deployed in hundreds, the chip technology was new, but unlike Apollo tech, it was ready for widespread deployment. The many uses of computers were well known even then, and physical limits were far away, so rapid growth was to be expected.

Actually computers had more or less hit the physical limits by then and were only saved by photolithography (the means to make a chip).  The issue was soldering just so many transistors to a board with zero failure.  This was called the "tyranny of the numbers" and it looked like the CDC7600 was simply the most power computer you could reasonably build (although I think the Cray 1 may have required primitive chips).  Photolithography meant you could have thousands (70s), millions (80s,90s), or billions (21st century) of active parts with zero failures* (of course chip size was limited.  There are bad chips, and the bigger the chip the more likely to include a fabrication error.  But if you have enough duplicate parts (say a memory, or a modern GPU) you can often only disable the tiny portion around the error and keep going.  This was difficult in the days of discrete transistors.

I think the "tyranny of the numbers" was more an IBM thing (they wanted to mass produce the things after all, and couldn't afford the techs needed to make zero defects possible), but it was definitely a thing until the chip showed up and saved the progress of computers.  Had the technologists been watching, they might have also noticed just how hard a fast moving industry can ram into a hard limit.

* The chip also has a curious property that makes all Denning principles useless because there is no way to tell a good chip from a bad chip until you finish the job and test it.  Oddly enough, the only place I learned about the tyranny of the numbers wasted far too many chapters on Denning principles (which were entirely orthogonal, but presumably the author didn't find out until writing all those chapters).  I can assure you that if you put thousands of transistors on boards to build a computer 1960s style the MTBF (mean time between failure) will quickly plunge to zero (worse if you use tubes).

[Guest]

In case it's not obvious, I was talking about the chip era. Transistor-based computers did hit their physical limits, so any discussion of growth in that time means talking about chips.

[tater]

Notice what NASA now considers possible commercial vehicles in this slide.

Falcon 9, FH, Starship (old pic), 2 and 3 stage NG, and SLS blocks 1/2. No, this doesn't mean ULA is not on the table, but interesting to see things open up.

[Silavite]

This is welcome news.

https://spacenews.com/nasa-seeks-a-rapid-launch-of-a-lunar-lander/

Quote

NASA is also likely to depart from a conventional cost-plus contract for the lunar lander. “We would, in essence, be buying a service to take our astronauts from the Gateway down to the moon and back,” said NASA Administrator Jim Bridenstine at the town hall. “We’re looking to those service providers to create the absolute best ideas that they have.”

 

[tater]

 

[WinkAllKerb'']

On 4/13/2019 at 5:10 AM, Bill Phil said:

Well...

 

[Guest]

I’m not sure... Trump apparently gave this directive to NASA and then soon after he was saying publicly how people shouldn’t be talking about the moon we should be aiming for Mars “which is part of the moon” -Trump

It’s hard to figure out what is going on right now, let alone what’s going to happen in 5 years... I do feel that using the moon as a proving ground and a place to develop our space capabilities is a good idea since if ANYTHING goes wrong on Mars any help is probably at least 3-4 years away from getting there assuming there isn’t hardware waiting to go when the call comes in, and depending on launch windows etc... so I’m definitely a fan of building up some permanent capabilities on the moon just because we need more capabilities! And a more matured industry servicing space outside of LEO. Musk seems to be looking to prove his Starship on the moon too (I.e. the Dear Moon mission and potential landings judging by the promotional materials) So yes to the moon! But the when is anybody’s guess.

Edited June 15, 2019 by Guest