Why it Was Almost Impossible to Put a Computer in Space

Linus Tech Tips
28 May 202417:19

Summary

TLDRThe video explores the challenges and innovations of deploying computers in space, focusing on the Spaceborne project. It delves into how NASA, HP Enterprise, and Kokia collaborated to create the first Edge Computing server for the ISS. The project faced hurdles like radiation and power constraints, leading to the development of Spaceborne 2, designed for practical applications like AI-powered data analysis. The video highlights the importance of edge computing for astronaut safety, showcasing how it reduces data transfer times from days to seconds. It also covers the technical specifications, cooling systems, and networking solutions tailored for space, emphasizing the balance between performance, reliability, and power efficiency.

Takeaways

  • 🚀 The International Space Station (ISS) is home to numerous computers, and the harsh conditions of space, including microgravity and extreme power and cooling constraints, make it challenging to operate standard computers.
  • 🔧 In 2017, a collaborative effort between NASA, HP Enterprise, and Kokia led to the creation of the first Edge Computing server, Spaceborne 1, designed to operate on the ISS for extended periods, though it faced initial issues with supercapacitors.
  • 💡 Despite the initial challenges, the project was considered a success, and in 2021, Spaceborne 2 was launched to explore practical applications for on-station computing, particularly in data analysis using AI.
  • 🛠️ The design of the Spaceborne 2 server includes a balance between storage capacity and power efficiency, with a focus on meeting the unique requirements of the ISS environment, such as the need for redundancy and power conservation.
  • 🔌 The Spaceborne 2 server features more storage than ever, with over 130 TB, a significant achievement considering the design challenges and the need for efficient use of the limited power budget.
  • 🛰️ The server's storage configurations were carefully chosen to balance performance, reliability, and power efficiency, with a focus on using SAS drives instead of NVMe for high-speed bulk storage due to their balance of these factors.
  • 🌐 The networking of the ISS presents unique challenges, with the Spaceborne 2 servers not having a normal internet connection but instead a private link back to Earth, limited to 1 megabit per second, and with periods of downtime to prioritize power generation.
  • 🔗 The servers are designed to be installed and managed by the ISS crew, with a focus on user-friendliness and safety, including a white-glove test to ensure no injuries from snagging or tearing of equipment.
  • 🔬 The Spaceborne 2 project has provided valuable insights into the practical applications of Edge Computing in space, including the use of AI for data analysis and the importance of redundancy and power efficiency in space environments.
  • 🤝 The project highlights the importance of partnerships and collaboration in advancing technology for space exploration and research, with Kokia's sponsorship and support being a key factor in the success of the Spaceborne 2 project.

Q & A

  • What are the main challenges in putting a computer in space?

    -The main challenges include the harsh radiation, microgravity, extreme constraints on power and cooling, and the need for reliable and efficient systems that can withstand the conditions of space.

  • How do space computers handle the harsh conditions of space?

    -Space computers are designed to handle these conditions by being replaced frequently, using redundant systems, and incorporating technologies that are specifically designed to withstand the rigors of space, such as the Spaceborne 1 and Spaceborne 2 servers.

  • What was the purpose of the first Spaceborne server, and what were the initial issues encountered?

    -The first Spaceborne server was designed to run on the International Space Station for an extended period. The initial issues included the supercapacitors in the SSDs being prone to radiation-related failures, which were later addressed in the second version.

  • How has the second version of the Spaceborne server improved upon its predecessor?

    -The second version of the Spaceborne server has moved beyond proof of concept and explores practical applications for on-station compute, especially for data analysis using AI. It also features more storage capacity and has been designed to handle the increased demands of scientific applications and data analysis.

  • What are the design challenges faced when putting a server in the ISS?

    -Design challenges include fitting the server into the limited space of the ISS, ensuring power and cooling efficiency, managing the network bandwidth, and ensuring the server can withstand the launch conditions and the harsh environment of space.

  • Why were supercapacitors not suitable for the Spaceborne 1 server?

    -Supercapacitors were not suitable for the Spaceborne 1 server because they were prone to radiation-related failures, which is a significant concern in the space environment where radiation exposure can be high.

  • How does the Spaceborne 2 server contribute to astronaut safety on spacewalks?

    -The Spaceborne 2 server contributes to astronaut safety by processing images of space suit inspections quickly, which helps to identify any potential hazards such as scratches or tears, thereby improving safety during spacewalks.

  • What are the power and cooling challenges faced by the servers in the ISS?

    -The power and cooling challenges include operating within a power budget that is less than a typical gaming rig, managing the heat generated by the servers, and ensuring that the cooling systems can handle the high demands of the space environment.

  • How do the servers in the ISS manage data transfer and communication with Earth?

    -The servers in the ISS manage data transfer and communication by using a private link back to Earth, which is limited to a speed of 1 megabit per second. They also buffer their communications to ensure that if a ping is successful, they can send data and hope that the connection stayed up during that time.

  • What are the storage solutions chosen for the Spaceborne 2 server, and why?

    -For the Spaceborne 2 server, the storage solutions chosen include SAS interface drives for high-speed bulk storage and a configuration with two drives operating in a data redundant mirror, providing redundancy and reliability in the event of a failure.

  • How does the network configuration of the Spaceborne 2 server support the needs of the ISS?

    -The network configuration of the Spaceborne 2 server supports the needs of the ISS by connecting both ISS's internal gigabit networks to a redundant switch inside the locker and providing a link between the two lockers at 10 gigabit, which is suitable for backups or multinode workloads.

Outlines

00:00

🛠️ The Design and Selection of Space Hardware

This paragraph discusses the harsh conditions of space, such as radiation, microgravity, and limited power and cooling, which make it difficult to operate computers. It highlights the issue of computers being sent to the International Space Station (ISS) being prone to failure due to these constraints. The paragraph then introduces the concept of Edge Computing, a type of computing that brings data processing closer to the source of data, to address these challenges. The first attempt at this was the creation of a spaceborne server by NASA, HP, and Kokia, which faced initial failures but was later considered a success. The second attempt, Spaceborne 2, aimed to explore practical applications for Edge Computing, particularly in data analysis using AI. The paragraph also touches on the importance of Edge Computing in improving astronaut safety by analyzing space suit integrity through machine learning, despite the challenges of data transfer time.

Mindmap

Keywords

💡Edge Computing

Edge Computing refers to the concept of processing data closer to the source of data generation, rather than in a centralized data-processing warehouse. In the context of the video, Edge Computing is essential for the International Space Station (ISS) to perform data analysis and processing quickly and efficiently. The script mentions 'Spaceborne,' the first Edge Computing server designed for extended operation on the ISS, highlighting the practical applications of on-site compute, especially for data analysis using AI.

💡International Space Station (ISS)

The International Space Station (ISS) is a space environment research laboratory that orbits Earth. It is a collaborative project among multiple countries and serves as a hub for scientific research, space exploration, and technology development. The video discusses the challenges of operating computers within the ISS environment, such as radiation, microgravity, and power constraints, and how Edge Computing servers like 'Spaceborne' are adapted to function in these conditions.

💡Radiation

Radiation in the context of the video refers to the high-energy particles and electromagnetic waves present in space that can cause damage to electronic components, including computers. The script explains that radiation-related failure was an issue with the supercapacitors in the SSDs of the initial 'Spaceborne' server, illustrating the harsh environment that computer systems on the ISS must endure.

💡SSDs (Solid-State Drives)

SSDs, or Solid-State Drives, are data storage devices that use NAND-based flash memory to store data persistently. They are known for their fast data access speeds and lack of moving parts. The video script mentions that the supercapacitors in the SSDs of the 'Spaceborne' server were prone to failure due to radiation, indicating the need for specialized components that can withstand the space environment.

💡AI (Artificial Intelligence)

Artificial Intelligence (AI) is the simulation of human intelligence in machines that are programmed to think like humans and mimic their actions. In the video, AI is highlighted as a key technology for practical applications on the ISS, particularly for data analysis. The 'Spaceborne 2' server is mentioned as being designed to explore these applications, emphasizing the role of AI in enhancing the efficiency of operations in space.

💡EVA Gloves

EVA Gloves are the specialized gloves worn by astronauts during Extravehicular Activities (spacewalks). The script discusses how these gloves were responsible for a significant number of space-suit injuries, and how the 'Spaceborne' server can be used to analyze photographs of the gloves for integrity, thereby contributing to astronaut safety.

💡Data Transfer

Data Transfer refers to the movement of data from one location to another, such as from the ISS back to Earth. The video script points out the inefficiency of traditional data transfer methods, which could take up to five days. With the 'Spaceborne' server, data transfer times are significantly reduced to just 45 seconds, highlighting the benefits of Edge Computing in space.

💡Power Constraints

Power Constraints refer to the limitations on the amount of electrical power available for use. In the context of the video, power constraints are a significant challenge for operating computers on the ISS. The script discusses how the 'Spaceborne 2' server and its components were designed to operate within the limited power budget of the ISS, showcasing the engineering considerations required for space technology.

💡Storage

Storage in the video refers to the physical components used to store data, such as hard drives or SSDs. The script mentions that the 'Spaceborne 2' server features over 130 TB of storage, which is an impressive feat given the design challenges and the need for reliability and power efficiency in the space environment.

💡Networking

Networking in the context of the video pertains to the systems and hardware used to connect computers and transfer data. The script discusses the unique networking challenges on the ISS, including the use of specialized connectors and the limitations of the station's internet connection. It also highlights the importance of redundancy and high-speed connections for the 'Spaceborne' servers.

💡Kodiak

Kodiak, mentioned in the video, is likely a reference to a sponsor or partner involved in the 'Spaceborne' project. The script discusses how Kodiak sponsored the storage solutions for the 'Spaceborne 2' server, emphasizing the collaboration between different entities in making space technology a reality.

Highlights

Space is a harsh environment for computers due to radiation, microgravity, and power/cooling constraints.

The International Space Station (ISS) regularly receives shipments of laptops that are often destroyed.

In 2017, NASA, HP Enterprise, and Kokia collaborated to create SpaceBorne-1, the first Edge Computing server for the ISS.

SpaceBorne-1's SSDs were found to be prone to radiation-related failures.

SpaceBorne-2 was launched in 2021 to explore practical applications of on-orbit compute, especially AI-based data analysis.

SpaceBorne-2 features over 130 TB of storage, a significant increase from its predecessor.

Edge Computing on the ISS can reduce data transfer times from days to seconds, improving astronaut safety and efficiency.

Initial skepticism about computers on the ISS was due to the harsh environment and the need for reliable, durable equipment.

Equipment for the ISS undergoes rigorous testing, including launch simulations and acoustic chamber tests.

The 'white glove test' ensures that equipment will not cause injury to astronauts.

The SpaceBorne servers are standard HP Enterprise systems without special armor or modifications.

Kokia sponsored the storage solutions for SpaceBorne, providing SSDs that are radiation-hardened.

The choice of SAS interface drives over NVMe was due to a balance of performance, reliability, and power efficiency.

The DL360 server has a unique configuration with two warm spares in a four-drive array for redundancy.

The EL4000 blade chassis houses slide-in blade servers with integrated NVMe drives.

Lockers on the ISS present design challenges, requiring creative solutions for mounting and cooling.

The ISS uses a combination of air and water cooling systems to manage the heat generated by servers.

Networking on the ISS is limited by power budgets and the need for redundancy and high-speed connections.

The ISS has a private link back to Earth with limited bandwidth and connectivity issues.

The SpaceBorne project aims to learn from the challenges and innovations in deploying and operating servers in space.

Hardware and software RAID cards are being tested for their impact on mass, power consumption, and reliability.

Transcripts

00:00

between the radiation the micro gravity

00:03

and the extreme constraints on Power and

00:05

cooling space is just about the worst

00:09

possible place to put a computer can you

00:12

guys let me

00:14

in that's better now if you follow space

00:18

research and exploration you probably

00:19

know that space is just full of

00:21

computers like this one so how do they

00:25

do that well the short answer is by

00:28

replacing them a lot the ISS takes

00:31

regular shipments of dozens of laptops

00:34

at a time which get and this is great

00:35

this is a direct quote absolutely

00:39

destroyed but not every computer can be

00:42

disposable and in 2017 the madlads at

00:45

Nasa HP Enterprise and kokia who

00:49

sponsored this video collaborated to

00:51

create spaceborn one the first Edge

00:54

Computing server that was intended to

00:55

run for an extended period of time on

00:58

the International Space Station

01:00

of course it being their first attempt

01:02

some uh let's say learning took place

01:05

and it turned out that the super

01:07

capacitors in the ssds were prone to

01:09

radiation related failure who knew but

01:12

since all it was ever meant to do was

01:14

run benchmarks anyway the mission was

01:16

considered a huge success and in 2021

01:19

they launched spaceborn 2 whose purpose

01:22

was to move Beyond proof of concept and

01:24

explore practical applications for onst

01:27

compute especially data analysis using

01:30

AI but the story doesn't end there

01:33

behind me is new space born 2 for

01:36

administrative reasons it has the same

01:38

name and Core specs as last time but it

01:40

just took off earlier this year and it

01:43

features more storage than ever over 130

01:46

tabt which is an incredible feat when

01:48

you consider the design challenges I

01:50

mean where do they even install these

01:52

things look up oh right I guess they

01:56

don't really need a ladder do they no

01:58

but you might

02:04

[Music]

02:09

to illustrate why Edge Computing is

02:11

needed on the ISS let's look at a use

02:13

case that's focused on astronaut safety

02:16

these are the Eva gloves that the crew

02:18

members wear during space walks and

02:20

according to this article from 2016 they

02:23

were responsible for half of all space

02:26

suit injuries so to ensure their

02:28

integrity between you es NASA requires

02:31

the crew to take hundreds of photographs

02:33

of them from every angle and then beam

02:35

them back to Earth where machine

02:37

learning is used to analyze them for

02:39

scratches or other hazards except for

02:41

one small problem that data transfer

02:44

takes five days but with spaceborne oh I

02:49

don't know how about 45 seconds not only

02:52

is this a huge time sa but with only a

02:55

handful of pictures needing to go to

02:57

Earth for further analysis spaceborn can

02:59

free up a significant amount of the

03:01

crew's limited network bandwidth for

03:03

other more interesting things with such

03:07

obvious benefits then you got to be

03:08

wondering why did no one ever try to put

03:10

a server on the ISS before the short

03:13

answer is after seeing how the crew

03:15

laptops fared many people thought that

03:17

they just plane wouldn't work and even

03:20

if they did there were a host of other

03:22

hurdles to clear like the launch okay

03:25

this is really cool rocket companies

03:27

like SpaceX and Northrop Grumman have

03:29

Shak test machines that are programmed

03:31

with profiles that will simulate the

03:33

launch conditions of their respective

03:35

rockets and if you've seen that viral

03:38

video of the machine that disassembles

03:40

hard drives by vigorously shaking them

03:42

you're going to know that surviving that

03:44

kind of treatment is no mean feed well

03:47

these machines Managed IT both in the

03:49

simulation and in the real world they

03:52

actually lift it off at the end of this

03:54

January every piece of equipment must

03:56

also pass an acoustic chamber test and a

03:59

US user friendliness evaluation to

04:01

ensure the station crew can install and

04:03

manage it and uh oh here's a good one

04:05

apparently all equipment sent up to the

04:07

ISS goes through what's called a white

04:10

glove test which uh thankfully is not

04:13

what it sounds like basically you put on

04:15

a pair of white gloves and then you just

04:17

manhandle the crap out of it if the

04:20

gloves snag or tear on

04:23

anything yeah that's a potential source

04:26

of injury I'm going to need you to file

04:28

that down which f fun fact they actually

04:30

do onsite and then repeat the test I

04:33

just hope they weren't filing any RAM

04:36

sticks speaking of let's take a closer

04:38

look at these machines machine Z cuz

04:41

it's not just one in here I know it was

04:44

kind of the point of this whole

04:45

experiment but it still weirds me out

04:47

that these are just bog standard HP

04:49

Enterprise systems that you could order

04:50

on their site today they don't even have

04:52

lead armor or anything in this case

04:54

we're looking at an edgeline 4000 which

04:57

is a multi-blade system and a dl360 dual

05:00

socket server we asked why these

05:02

specific machines and the answer we got

05:04

was shockingly relatable we sorted the

05:07

HP server catalog by depth power draw

05:10

and GPU support and these were the ones

05:12

we were left with all right fair enough

05:16

as for why two different machines well

05:18

here's the thing in a perfect world

05:21

multiples of the same machine would have

05:23

been better but due to power constraints

05:26

they chose to have one with more CPU

05:28

cores for more additional scientific

05:30

applications and one with pure CPU cores

05:34

but with a GPU for deep learning and AI

05:37

one thing they needed for both however

05:38

is Ample Storage kokia generously

05:41

sponsored this and brought us out here

05:43

so let's take a look at the let's call

05:46

them unique choices that they made for

05:48

their storage configurations first up

05:51

obviously gone is any trace of super

05:53

capacitors so kokia can proudly say that

05:56

their ssds are space ready I guess but

05:59

but what's less obvious is why they

06:02

chose a SAS interface Drive rather than

06:05

nvme for their high-speed bulk storage I

06:08

mean you would think this is Space Age

06:10

Technology they'd want the fastest thing

06:12

possible but these drives were selected

06:15

for their balance of performance

06:17

reliability and especially power

06:20

efficiency when you're looking at a

06:22

shared power budget across two servers

06:25

that is less than a typical gaming rig

06:28

every single lot counts oh right and

06:32

that's even under ideal conditions at

06:34

any given time to conserve power for

06:37

other priorities on the station the team

06:39

can be asked to operate in half power

06:41

mode or even to shut down entirely for

06:44

large operations like docking so the new

06:46

DL 360 server this guy right here gets

06:49

four 3.72 TB pm6 Enterprise drives

06:54

totaling

06:55

120 tabt of raw bulk storage for

06:58

scientific data and for backups then for

07:02

application drives we've got a really

07:04

wild config again they went with four

07:06

drives but this time it's their RM sixs

07:11

again they're using SAS for lower power

07:13

but this time two of the drives are

07:15

operating in a data redundant mirror and

07:18

the other two are basically just

07:20

chilling there ready to be put into

07:21

action in the event of a failure two

07:24

warm spares out of a four Drive array

07:27

would sound like crazy Paran oia on

07:30

Earth but I assure you that in space

07:32

where bit flips from random radiation

07:35

are much more common it's perfectly

07:37

reasonable I mean other than the

07:39

overkill Drive config and the 28v power

07:43

conversion that they need to run it on

07:45

the ISS there's not much to say about

07:47

this thing it's pretty much a bog

07:49

standard server there is one cool demo

07:51

that they said we could run though oh

07:53

yeah they offered to let us pull one of

07:55

the drives out of this dummy machine and

07:58

live swap it in into this running

08:00

machine to show that no data loss will

08:02

occur you want to do the honors sure all

08:05

they asked is that you put it in Bay 8

08:07

see you can see the drive is actually

08:08

operating use Bay s that works look at

08:11

that I mean that's good rm6 that would

08:13

be bad if it wasn't

08:15

kokia and then let's check the size

08:18

beautiful 3.8 tab exactly what we want

08:22

they're not the exact drives that are on

08:24

the space station but we wanted a

08:25

different capacity to show you that it's

08:26

working and those are expensive

08:30

look at that status rebuilding I mean

08:32

this seems like a lot of extra steps we

08:34

could have just looked up the light yeah

08:35

the light yeah it's going

08:38

success and whether you're looking for a

08:40

SAS drive an nvme drive high capacity or

08:43

high performance we're going to have a

08:44

bunch of Kyo has great Enterprise grade

08:47

drives Linked In the description down

08:49

below I think I'll let you take this one

08:51

apart it's uh appropriately line this

08:53

size okay let's take a look at the

08:55

second server that's packed into each

08:57

Locker the el4 ,000 is a blade chassis

09:01

so the servers are basically these

09:04

little slide in cards that yeah I know

09:07

AR these cute or what wait pull it out

09:09

and they go in on the side look at that

09:12

instead of from the front wow and they

09:15

managed to sneak four kokia xg6 nvme

09:20

drives into each of these blades well

09:24

when I say each of these blades I should

09:26

say they had the power budget for four

09:28

drives but they didn't have the power

09:30

budget for four blades in the flight

09:33

configuration of this system they ship

09:35

with just one of the four blades

09:38

installed though it should be noted they

09:40

do fly up a spare blade per system in

09:42

the event of a failure there's just no

09:45

way that that poor Locker can support

09:47

both these blades and the other server

09:49

running concurrently let's put you away

09:52

and shift our Focus to the locker now

09:55

obviously there's no real up or down on

09:59

the

09:59

as it whizzes around the earth at around

10:01

28,000 km an hour but to improve comfort

10:05

for the astronauts they tend to mount

10:06

directional items like plants in a fixed

10:09

orientation which will put our lockers

10:12

the drawers that hold our servers in the

10:15

ceiling there are two of these lockers

10:18

each containing an identical system load

10:20

out for workload sharing and redundancy

10:22

and these lockers present some serious

10:24

design challenges starting with the fact

10:27

that they use a standard that quite lit

10:29

Lally doesn't exist on Earth Express

10:32

rack to pack the servers in then HP

10:35

Enterprise had to get kind of creative

10:37

they found the shortest servers they

10:39

could and then they stuffed them in

10:41

sideways and they're using a combination

10:43

of air Cooling and water cooling the air

10:46

cooling uses a system on the ISS called

10:48

AAA so at the back of these lockers

10:51

there's two cold air supplies and then

10:53

two hot air returns that handles about

10:56

20% of the cooling for the servers op

10:59

viously 20% not 100% they're going to

11:02

need some more and that's where this

11:04

water cooling comes in this isn't a one:

11:06

one for how it would be deployed on the

11:08

ISS for one thing these fittings 3D

11:11

printed mockups real fitting $800 a pop

11:15

if you were even allowed to buy them

11:18

this tubing cheap vinyl from Home Depot

11:21

real tubing must be made of stainless

11:23

steel in fact any wetted surface so

11:25

anything that comes in contact with

11:26

water is supposed to be made out of

11:28

stainless steel but we can still

11:30

illustrate how it's supposed to work so

11:33

on this side these go into a heat

11:35

exchanger much like this one this is

11:37

actually from the first generation

11:39

spaceborn but functionally it's the same

11:42

it pumps cold water into here chills the

11:46

air inside the system and then takes the

11:48

warm water out to be dissipated to the

11:51

photovoltaic heat exchangers that are

11:53

plumbed up with liquid ammonia coolant

11:55

and mounted to the exterior of the

11:56

station to sink that heat into space

11:59

you need these kinds of special heat

12:01

exchangers because while we think of

12:03

space as cold and we see people you know

12:07

oh blasted out of airlocks and they

12:10

freeze over or whatever in movies the

12:13

truth is that for traditional methods of

12:15

heat dissipation you need air and in the

12:18

near vacuum of space well it ain't

12:22

there air get

12:25

it cheesy jokes aside the two cooling

12:27

systems together are good for removing

12:29

about 400 watts of heat from each locker

12:32

but that's a combined budget so if this

12:35

GPU server kicks into high gear well

12:38

these CPUs better just chillax for a

12:41

little bit now let's talk about one of

12:43

my favorite subjects networking there's

12:45

four standard RJ45 ports on the front of

12:48

the okay I'm going to show you on the

12:50

real one ah as I was saying four ports

12:53

on each Locker two of them connect both

12:56

of the iss's internal gigabit networks

12:58

to a separate redundant switch inside

13:01

the locker and then the other two links

13:03

are going to go between the two lockers

13:05

at 10 GB why 10 gig well because for

13:09

either backups or for multinode

13:11

workloads that is a heck of a lot better

13:13

than gigabit and the power budget didn't

13:16

allow for anything faster cool I guess

13:18

I'm starting to notice a pattern here

13:21

anyway that's all pretty standard but

13:24

things become less so when you look at

13:26

the station side of these Networks

13:29

cables this is a 37 pin military spec

13:33

locking connector these are designed for

13:36

power and data but in this application

13:38

just eight of the pins would be used and

13:41

it is

13:43

$220 for just this part now on the space

13:46

station NASA provides these cables for

13:49

you but for testing sake really here on

13:51

Earth HP Enterprise had to make their

13:54

own

13:56

fantastic what's really going to blow

13:58

your mind though is for all of their

13:59

expensive networking these machines do

14:02

not have a normal internet connection

14:04

just a private link back to Earth that

14:07

NASA not only limits to a mighty 1

14:11

megabit per second but that they also

14:14

encourage folks not to make full use of

14:17

also even now in 2024 it doesn't have

14:20

247 connectivity pretty much every hour

14:24

or two there's a period of downtime that

14:26

can be anywhere as short as a few

14:27

minutes or as long long as 45 minutes

14:31

and that's because they have to

14:32

prioritize generating enough power for

14:34

the station and when the giant solar

14:36

arrays Point toward the sun they can

14:39

block line of sight with the satellites

14:41

that provide connectivity which oh

14:43

that's a fun fact even though the ISS

14:46

orbits less than 500 km from the surface

14:49

of the Earth our ping times to the ISS

14:52

and yes we got to Ping the ISS which was

14:56

pretty cool but our ping times were a

14:59

atrocious reaching nearly a second as we

15:02

uploaded some of the dankest memes that

15:04

Earth had to offer LT store.com now we

15:08

asked why that is and the answer was

15:10

twofold one it's really old okay fair

15:15

enough but also too the station's

15:18

Internet relay is in geosynchronous

15:20

orbit over

15:22

35,000 kilm from the Earth's surface and

15:25

uh well what I said was H well there's

15:28

your problem right there it's just

15:30

really far and you might be wondering

15:32

well why don't they just use

15:34

starlink that's a good question um

15:36

someday they might but for now they

15:38

don't and HP and the team on the ISS

15:41

have to work around the constraints of

15:42

the current setup I mean for crying out

15:44

loud it took them four years to validate

15:46

that you know we can even just run a

15:48

normal computer up here and actually

15:49

expect this thing to be reliable they

15:51

can't just switch to something and go I

15:53

don't know I hope it works um oh by the

15:55

way here's another fun one there's no

15:57

API to determine if if their connection

15:59

is up or down so instead what they do is

16:02

buffer all their Communications in basic

16:05

terms that means that they ping every

16:07

second and if the ping succeeds they

16:10

send data and then hope that the

16:12

connection stayed up during that time

16:14

it's a pretty good system okay not a

16:17

perfect one but certainly enough for us

16:19

to learn a lot from the spaceborn

16:21

project even though spaceborn 2 has been

16:24

in action for 3 years there's still so

16:26

much to learn two of the servers have

16:28

hard Hardware raid cards for their

16:30

drives for example costing both mass and

16:33

power consumption while two of them use

16:35

software raid which obviously doesn't

16:37

consume any Mass but could impact power

16:40

consumption even more depending on the

16:42

loads and one of them could be more or

16:45

less reliable than the other we won't

16:47

know until we try which is kind of a

16:50

recurring theme here so if you want to

16:52

learn more about the spaceborn computer

16:54

project we're going to have some

16:55

resources linked for you down below and

16:57

we're also going to have a L to some

16:59

great Enterprise storage options from

17:01

our sponsor kokia we're truly grateful

17:04

for this unique opportunity to get

17:06

realistically as close as I ever will to

17:08

the real ISS a prop in a sound stage in

17:13

LA but hey thanks kogia for the

17:15

opportunity and for your long-term

17:17

partnership

Rate This

5.0 / 5 (0 votes)

Связанные теги
Space ComputingEdge ServerISSNASAHP EnterpriseKokiaAI Data AnalysisHeat ManagementNetworkingStorage SolutionsSpace Technology
Вам нужно реферат на русском языке?