NASA and HP launching HAL to ISS?

[WMGroomAK]

NASA is working together with Hewlett-Packard Enterprises (HPE) to launch a 1-teraflop computer to the International Space Station (ISS) to study the effects of space on higher power computers as part of their research into future deep space travel.  As it stands the ISS is running on i386 processors and has a lot of support from Ground Control, however, NASA figures the latency from deep space travel will make this current method impractical and those missions will require more powerful computers, potentially with AI, in order to properly maintain the spaceships systems and navigation.

https://arstechnica.com/science/2017/08/spacex-is-launching-a-supercomputer-to-the-international-space-station/

Quote

One of the critical technologies NASA will need if it really does send humans beyond the Earth-Moon system within the next few decades is more powerful computers capable of operating in the deep space environment. Presently, the main command computers that operate the space station use Intel i386 processors. However, that is fine for the station because all of its critical systems are monitored around the clock by ground-based flight controllers who can work in real time with the crew to fix any problems that arise.

If humans do travel to Mars, they will face increasingly long communications delays—stretching out to more than half an hour—between Earth and their spacecraft. In that situation, the astronauts are likely to become more reliant on more powerful computers and artificial intelligence to make critical course corrections or decisions within seconds or minutes.

 

A "smart" spacecraft, however, will require a considerably more powerful and robust computer. So NASA and Hewlett-Packard Enterprise (HPE) are taking the first step toward that by launching a "supercomputer" to the International Space Station. It will ride into space as early as Monday aboard SpaceX's next supply mission to the station.

...

For the year-long experiment, astronauts will install the computer inside a rack in the Destiny module of the space station. It is about the size of two pizza boxes stuck together. And while the device is not exactly a state-of-the-art supercomputer—it has a computing speed of about 1 teraflop—it is the most powerful computer sent into space. Unlike most computers, it has not been hardened for the radiation environment aboard the space station. The goal is to better understand how the space environment will degrade the performance of an off-the-shelf computer.

 

During the next year, the spaceborne computer will continuously run through a set of computing benchmarks to determine its performance over time. Meanwhile, on the ground, an identical copy of the computer will run in a lab as a control.

Any chance we can get @LinusTech or @Slick to catch a ride on this rocket and provide the first LMG video in SPACE!?  Or even maybe just get a hold of NASA and get a tour of their ground based control?

 

Official NASA Experiment Page: https://www.nasa.gov/mission_pages/station/research/experiments/2304.html#overview

 

Quote

Experiment Description
Research Overview

• Computer and data intensive applications are run on the Spaceborne Computer systems.
• The power consumption is monitored and power usage is dynamically tuned during these runs.
• The effects of radiation on the systems when running performance applications are determined concurrently with detecting/analyzing/adapting to data, quickpath interconnectTM (QPI) internal and FDR external errors.

Description
The research objectives of the Spaceborne Computer include a year-long experiment of operating high performance commercial off-the-shelf (COTS) computer systems on the ISS with its changing radiation climate. During high radiation events, the electrical power consumption and, therefore, the operating speeds of the computer systems are lowered in an attempt to determine if such systems can still operate correctly. Additionally, this is a long duration experiment, studying the practicality of running and managing COTS high performance computer systems in orbit from several months to one year. In summary, the objectives are:  1) run compute and data intensive applications in a changing radiation climate, 2) monitor power consumption and dynamically tune the power consumed, and 3) determine effects of solar radiation on the systems while running. In order to achieve these objectives, Hewlett Packard Enterprise (HPE) proposes a total of four identical high performance COTS computer systems. Two of the systems are installed aboard the ISS in a side-by-side EXPRESS locker within an ISS EXPRESS Rack. These two systems with the required networking are integrated at the HPE facility and turned over to the ISS Cargo Mission Contract (CMC) as required.

 

Edited August 11, 2017 by WMGroomAK
Additional Article and info

[bcredeur97]

I could see us having high powered computers orbiting the earth (or even other planets too) and using the advantage of those being already in space to more easily transmit data to missions throughout the solar system (no atmosphere for signals to pass through probably helps a lot if I had to guess)

 

and then they could all be linked together and all the data could be accessed from any planet or system.

[Zando_]

2 minutes ago, WMGroomAK said:

Any chance we can get @LinusTech or @Slick to catch a ride on this rocket and provide the first LMG video in SPACE!?  Or even maybe just get a hold of NASA and get a tour of their ground based control?

THAT.... WOULD BE AWESOME!!!

[Himommies]

2 minutes ago, WMGroomAK said:

i386

really?and then there just moving to 1 teraflop.shoudn't they take it a little more slowly

[Tech_Dreamer]

5 minutes ago, WMGroomAK said:

Any chance we can get @LinusTech or @Slick to catch a ride on this rocket and provide the first LMG video in SPACE!?  

imagine they dropping that shit from space all the way to the ground

[AHaskin14]

Never knew the majority of monitoring was handled solely from the ground (and I've been to the room where that's done). I assumed there was more self sustaining activity happening on board the station. Very cool to watch this develop.

[TacoSenpai]

8 minutes ago, WMGroomAK said:

Any chance we can get @LinusTech or @Slick to catch a ride on this rocket and provide the first LMG video in SPACE!?  Or even maybe just get a hold of NASA and get a tour of their ground based control?

Linus:  Can we get on the ISS to do the WAN show from space?

 

HAL:  I'm afraid I can't let you do that Linus...

[Flavio hc 16]

3 minutes ago, LordTaco42 said:

Linus:  Can we get on the ISS to do the WAN show from space?

 

HAL:  I'm afraid I can't let you do that Linus...

I'm not going to drop the ISS from space...maybe 

[Jito463]

I, for one, welcome our new AI overlords.

[Sauron]

15 minutes ago, bcredeur97 said:

I could see us having high powered computers orbiting the earth (or even other planets too) and using the advantage of those being already in space to more easily transmit data to missions throughout the solar system (no atmosphere for signals to pass through probably helps a lot if I had to guess)

 

and then they could all be linked together and all the data could be accessed from any planet or system.

Honestly sending data to Earth's surface from planetary distances is probably only marginally easier than transmitting to the surface. This probably helps the ISS do heavy calculations locally instead of having to beam down massive data sets, but I don't see it being particularly useful for other objects in orbit or otherwise.

[Guest]

Computers in Space!

 

We need Supercomputers in the bottom of the ocean too.

 

What happened to that Microsoft underwater datacenter project? Did it work?

[Curufinwe_wins]

So for people wondering.... There is not insignificant reason to suspect that newer (read smaller) processing nodes are more susceptible to deep space data/process corruption.

 

 

[Flavio hc 16]

5 minutes ago, Curufinwe_wins said:

So for people wondering.... There is not insignificant reason to suspect that newer (read smaller) processing nodes are more susceptible to deep space data/process corruption.

 

 

Of course smaller nodes are more susceptible to data corruption: the smaller and more complex a node is, the more likely radiation can cook it or render it unusable, especially because you render more likely quantum tunneling and such when elettrons get "excited " from cosmuc radiation.

[Vivace]

34 minutes ago, Goku-sama said:

What happened to that Microsoft underwater datacenter project? Did it work?

They actually deployed it. Project is still operating. Here's the official project site.

[ScratchCat]

11 minutes ago, Flavio hc 16 said:

Of course smaller nodes are more susceptible to data corruption: the smaller and more complex a node is, the more likely radiation can cook it or render it unusable, especially because you render more likely quantum tunneling and such when elettrons get "excited " from cosmuc radiation.

Well that's great, now I can't play CS:GO during WW3

I would have thought they would have advanced to a new node for their processors. i386s seem a bit slow for the biggest space station. I guess most data collected is processed on earth and doesn't influence the station and all course control can be computed by NASA in advanced

[Taf the Ghost]

That's a big jump, so maybe they should start with a Pentium II and we'll see from there?

 

There's a joke about "no wonder it has worked for so long, it wasn't running Windows 95!".

 

Space is a very different environment, it'll be interesting to see what they put up there. Granted, at least you really won't need much in the way of heatsinks.

[MoonSpot]

2 hours ago, WMGroomAK said:

Any chance we can get @LinusTech or @Slick to catch a ride on this rocket and provide the first LMG video in SPACE!?  Or even maybe just get a hold of NASA and get a tour of their ground based control?

Get a Co-op with Destin from Smarter Everyday on board and you'll have the ingredients for some really epic vids.  I'm actually drooling at the thought.

[agent_x007]

Problem with high density CPUs is complexity.
Less transistors = less things that can go down beacuse of space radiation, microasteroids, etc.
I'm conviced tho, that even with quadruple backups for any SoC component, modern tech could do better than i386 performance...

[Bananasplit_00]

2 hours ago, Curufinwe_wins said:

So for people wondering.... There is not insignificant reason to suspect that newer (read smaller) processing nodes are more susceptible to deep space data/process corruption.

 

 

Hmm makes me want to see both an I386 made on 10nm and a modern chip made on a couple hundred nm lithography... 

 

38 minutes ago, agent_x007 said:

Problem with high density CPUs is complexity.
Less transistors = less things that can go down beacuse of space radiation, microasteroids, etc.
I'm conviced tho, that even with quadruple backups for any SoC component, modern tech could do better than i386 performance...

I'm pretty sure an I386 would get rekt by a micro asteroid just as much as a 7700K

[SansVarnic]

I want to know how well that i386 stood up to the wear and tear of Space.  

Apparently very well if it is still running.

[HalGameGuru]

wait I'm not ready to go and I don't know how to lock the pod bay doors!

[Dabombinable]

386. The CPU that just keeps on going.

[Curufinwe_wins]

11 hours ago, Bananasplit_00 said:

Hmm makes me want to see both an I386 made on 10nm and a modern chip made on a couple hundred nm lithography... 

 

I'm pretty sure an I386 would get rekt by a micro asteroid just as much as a 7700K

A modern chip on anything above 28nm would be insanely expensive... cost per transistor continues to fall.

 

But the i386 made on something like 14nm would actually be interesting.

[Dabombinable]

12 minutes ago, Curufinwe_wins said:

A modern chip on anything above 28nm would be insanely expensive... cost per transistor continues to fall.

 

But the i386 made on something like 14nm would actually be interesting.

As it is they consume little power (relatively speaking) and don't require any active cooling (even for the AMD 386DX-40), so at 14nm they could probably be powered by the kind of solar cells found in calculators. Though the lack of an FPU might be an issue still since you'd also need to have a 14nm 387 co pro.

[matrix07012]