The Computer on the Space Shuttle Sucks

HardOCP News

[H] News
Joined
Dec 31, 1969
Messages
0
So I am reading this article about the delay in launching the space shuttle Atlantis and I noticed this:

Bottom line: If Atlantis still is on the ground Dec. 14, its launch would be delayed until around Jan. 7. NASA would avoid flying during the New Year's holiday because the shuttle's computers are not designed to handle the year-end rollover.

Wait. What?
 
"Houston, our only chance is to cancel the mission and crash land in the pacific!"
"Whats wrong? is the hull breached? are your fuel tanks leaking?"
"No, in excactly 1 hour, our windows 7 beta will expire"
 
Just remember when those systems were developed (early 70's), how rigorously they're tested, and how much bureaucratic red tape may be in the way of updating said systems (what with the orbiter sun-setting in the next few years). I think I read that the next OS will be UNIX-based, providing MUCH greater flexibility, and the hardware mostly off-the-shelf instead of almost entirely custom/proprietary. Ahh, the benefits of being designed/built after the computer industry's boom.
 
"Houston, our only chance is to cancel the mission and crash land in the pacific!"
"Whats wrong? is the hull breached? are your fuel tanks leaking?"
"No, in excactly 1 hour, our windows 7 beta will expire"

lol...ha....love it!
 
Oh come on Steve, you act like they have billions of dollars to spend in their space program that could get a computer system that could handle a task like that.

Oh..wait...nevermind...carry on.
 
yes nasa computers do suck i am shocked that they have not taken the time to build a NEW damn computer system for the damn thing hell remember Armageddon with the new space shuttles why cant they not just build a new updated one
 
What they use for computer hardware is absolutely archaiac by modern standards.
 
Oh come on Steve, you act like they have billions of dollars to spend in their space program that could get a computer system that could handle a task like that.

Oh..wait...nevermind...carry on.

You act like a consumergrade PC would survive in space, let me guess, you don't work in that industry and havn't read anything about it?

On a 45nm CPU all it takes is 1 charged alpha particle to flip a bit....not a good thing.
 
Just remember when those systems were developed (early 70's), how rigorously they're tested, and how much bureaucratic red tape may be in the way of updating said systems (what with the orbiter sun-setting in the next few years). I think I read that the next OS will be UNIX-based, providing MUCH greater flexibility, and the hardware mostly off-the-shelf instead of almost entirely custom/proprietary. Ahh, the benefits of being designed/built after the computer industry's boom.

Isn't it running Debian right now?
 
You act like a consumergrade PC would survive in space, let me guess, you don't work in that industry and havn't read anything about it?

On a 45nm CPU all it takes is 1 charged alpha particle to flip a bit....not a good thing.

psst, we're talking about software, not hardware.
 
You act like a consumergrade PC would survive in space, let me guess, you don't work in that industry and havn't read anything about it?

On a 45nm CPU all it takes is 1 charged alpha particle to flip a bit....not a good thing.

But think of the overclock potential in space.
 
You act like a consumergrade PC would survive in space, let me guess, you don't work in that industry and havn't read anything about it?

On a 45nm CPU all it takes is 1 charged alpha particle to flip a bit....not a good thing.

Wow, you actually went there. Do you think a bunch of guys who roam the [H] know stuff about charged alpha particles? Chill out, I think he was going for irony buddy.
 
Wow, you actually went there. Do you think a bunch of guys who roam the [H] know stuff about charged alpha particles? Chill out, I think he was going for irony buddy.

Isn't that where this WAS going? I mean, the computer for the space shuttle isn't your run of the mill x86 pc, nor is it the computer from the Enterprise either.
I agree that they should have developed a replacement more advanced than what they have now, but with the bureaucracy NASA has, that will be in the year 2106 or when warp drive is invented.
 
I honestly don't see how they could develop a full-proof avionics system that runs on linux or any open source project. The whole comment about the flipped particle was probably inferring that if you want to run standardized non-military software, you're going to be running it on some x86 processor that hasn't be thoroughly tested for harsh conditions that allow no room for error. It's not like you can just drop a core i7 into the space shuttle and boot everything up on linux. Accidentally fill up kernel memory and the shuttle crashes into the ocean. Yeah that's a good idea.
 
Wha wha what? You mean Apple hasn't conned, er convinced NASA to use Macs in the Space Shuttle? Come on, "Snow Tiger in Space" practically writes itself, don't cha know?
 
IIRC the International Space Station runs x86, specifically radiation proofed 386's. I remember the article being printed after the pentium came out and NASA wasn't going to use the new chip because they had already spent 20-30k shielding each 386.

Of course, on a recent video walkthrough of the station I saw there were laptops everywhere, so who knows what they've got going now....
 
You act like a consumergrade PC would survive in space, let me guess, you don't work in that industry and havn't read anything about it?

On a 45nm CPU all it takes is 1 charged alpha particle to flip a bit....not a good thing.

Saying "charged alpha particle" is like saying wet water. Alpha particles (helium nuclei) cannot be "un-charged" ;)
And consumer grade cpus could survive in space, if you have a few of them working in parallel to verify results.

Anyway, this thread is about idiotic software that can't survive the year change however fast the CPU it is running on...
 
I bet the space shuttle computers would be just fine, but this isn't a TESTED SCENARIO. Can't intentionally fly in an untested scenario. Simple as that.
 
Saying "charged alpha particle" is like saying wet water. Alpha particles (helium nuclei) cannot be "un-charged" ;)
And consumer grade cpus could survive in space, if you have a few of them working in parallel to verify results.

Anyway, this thread is about idiotic software that can't survive the year change however fast the CPU it is running on...

ever heard of dry ice? :-P
 
SSbsod.jpg


Ouch, that would suck for sure.
 
OK.. color me confused. What does it matter if the thing thinks it is 1910 or 2010? It will still fly and do all the other stuff, as for records, we obviously know that there was not a space shuttle in 1910.
 
wow. 386's

386's would be a tremendous upgrade to the shuttle. They currently run five IBM AP-101S mainframe computers (four redundant primary, one backup) with about 1 Meg of memory and about 1.2 MIPs. Changing those out requires new software, OS, flight cerification (huge expense), etc. The shuttles have exceeded their expected lifespan, so nobody wants to invest in new hardware/software/certification for just one or two more missions per shuttle. Not cost effective when there's something that already works well enough.

Space is a very harsh environment and it costs a great deal to get anything up there, so only tested, ruggedized, validated components are used, perferably MILSPEC or better. The last standard CPU to be specifically designed and tested for space is a special version of the 386 with special radiation sheilded packaging.
 
wow. 386's

Yes, and that's a speed demon compared to the space shuttle. The AP101S only does around 480,000 instructions per second. That's around the performance of a 8080. And keep in mind that this isn't the original processor as they upgraded it around 1990.
 
Honestly NASA needs to be scrapped, it runs like any other bloated government agency now. We would all be better off if all that budget were given to a new agency ran by JAXA .
 
Saying "charged alpha particle" is like saying wet water. Alpha particles (helium nuclei) cannot be "un-charged" ;)
Not to mention an alpha particle wouldn't penetrate the cpu casing, let alone the hull of the shuttle. ;)
 
OK.. color me confused. What does it matter if the thing thinks it is 1910 or 2010? It will still fly and do all the other stuff, as for records, we obviously know that there was not a space shuttle in 1910.

I doubt that would be the issue. The issue is the software isn't certified to be operating in-flight during a year change (any year change). Not being certified, and straight up not working are not necessarily the same, but it must be certified to do what they actually go out and do, and there probably isn't enough time between now and then to actually certify it.
 
Honestly NASA needs to be scrapped, it runs like any other bloated government agency now. We would all be better off if all that budget were given to a new agency ran by JAXA .

no what they need are competent administrators
 
Shuttles are like laptops, it's difficult to upgrade one... gotta wait for the new model.
 
NASA just isn't the cutting edge science agency it was in the 60s. Neither is Bell Labs. We need a new challenge for a new space program, for example, to develop technology that will take man out of this solar system in the next 20 years. The tech does not exist today, but if we put up the effort to do this seriously, it can be done.

Everything we have today is not new, it has been leveraged off research done 50 years ago. All we have done is take the same tech and miniaturize it. The only advances of note would be in biology and materials. We need to do away with combustion based propulsion systems for one thing.
 
What they use for computer hardware is absolutely archaiac by modern standards.

There is a reason for this. Computers and storage mediums that are sent into space, have to be hardened to work in space. Standard commercial parts would not survive being exposed to radiation of space (the Van Allen Radiation belt for example).
 
Back
Top