SpaceTEE - out of this world programmable cryptography

Transcript

Okay. So, I'm Daniel um co-founder of the space computer project and uh we'll take you through a interesting journey. There will be some um some aspect of the motivation for what we actually are building and then uh we'll also get into the deep dive of the technical details of how we're actually implementing what the space computer is. So a brief history we started with the big bang some billion years ago or something and since then there was the chaotic planetary scale toe vavo and uh all kind of um organisms gradually formed initially I think algae and and

some other organism eventually crawled out of the water. Um fast forward a little bit we became apes. I know some people still refer to us as apes. Um, from that the the hunter gatherers that already started being a little bit more strategic and then the early cavemen and the first formation of villages and essentially the kind of buds of human society and civilization. And then with that we were able to also get to the very early version of money with shells and rocks as medium of exchange and store of value.

and eventually establishing the global trade routes. So that would have been the silk road or the ships that sail on the sea otherwise with trading outputs out outposts here and there. Um the invention of the banking system, then the stock exchanges, venture capital systems, all of that that actually like still runs today. it's uh already existed for a while and maybe taking a tangent on a a different evolution. So if we're looking at compute starting from concepts that were initially um philosophical or the initial Greek math and then taking it

into a little bit more pragmatic and and useful forms beyond only um things that we can do ad hoc. We came up with task specific computers. task specific. So they were basically fit for certain unique applications but not for any application. Then came the brilliant Alan Touring that I'm sure many of you know. Alan Touring was the invention of the touring complete a touring computer machine which essentially is a general purpose computer. The motivation back then was to crack the enigma code and uh to defeat Nazi Germany in the second world war. But the interesting thing is that beyond only um solving that one problem,

Touring was a nerd and as a nerd his motivation was to bring complete new innovation which was a general purpose compute machine like he did not predict that we'll end up with applications like uh Google or Uber or Ethereum or Bitcoin running on a global scale computer and then to scale the computer or computer systems um the invention of the solid state transistor came. So essentially moving from mechanical moving parts to um solid state siliconbased chips that were then uh mini miniaturaturized and like to this day we're just like keep grinding at developing uh new nodes semiconductor industry and um that's

kind of how we scale compute. And then um as things advanced we came with the high level programming languages and then uh normal people could start writing scripts often time in a human readable language and they don't need to necessarily know all the machine level uh operations that actually happen underlying silicon chips and eventually graphic user interfaces which is you know um user friendly for the ape level which is going back to where we started. And now we take a little bit of a zoom in on the internet. So if we started with the web 1.0 in the early days that the 90s or so that it became a household thing, we had the Netscape and MIRC and the things that

were still relatively peer-to-peer. There was a little bit less conversation about like oh like engagement optimization, call to action and all the data harvesting that later came with web 2.0. So web 2.0 we moved from a more emergent and peer-to-peer like to a more user server uh architecture. So that would be you know all the centralized big tech companies that give fantastic service and commercialize it. But then at the same time we are quite well aware of the trade-offs that that we have with privacy breaches and our data basically being monetized by uh somebody else and we don't see so much of it on our own. and then the

introduction or the evolution into the web 3 ecosystem which is kind of a return or peer-to-peer variant of um of web 2 in a way and taking a little bit of a deeper look into it. So if we think about um civilizational technology and the place for crypto as a civilizational technology, the first breakthrough was the Bitcoin blockchain and Bitcoin was intended as peer-to-peer electronic cash system. also a payment system which ended up not really functioning that way for various reasons that I'm sure some of you are aware some less but um more landed on the use case of um digital gold and the bitcoin blockchain was somewhat akin to a task specific

distributed computer just like before the general purpose computer and that's that's a part of what the innovation of Ethereum was that it essentially expanded the uh design space with uh the introduction of the world computer essentially a general purpose smart contract platform um and and that enabled all the many use cases that we are thinking of today DeFi and Dows and NFTs and um memecoins and essentially you know open canvas but if we think about it since Ethereum there hasn't been something that was fundamentally new there was always like this is ch cheap cheaper or faster or uh something else like no offense for

Solana friends and I really respect but like if you actually think about it there's really no novelty since Ethereum so the question is is that is that what we are all about like just kind of a dot in the universe and like is there a place here to expand and now we'll go to another little uh tour with the space frontier so space has been hot for a while then dwindled and Now we've been really entering a reacelerated space phase and what we're seeing is that uh in the last two years alone there's been more launches of satellites into orbit than in the um previous century combined like the the

whole moon missions in the cold war etc. Right? That was kind of minuscule number compared to what we see launch into orbit nowadays. And we're seeing um innovations like the upcoming uh Starship and other competitors that are looking to increase the volume of the payload to hundreds of magnitude basically um compared to what we have today. So essentially we see that there is a new golden age of space and we can start thinking about new applications beyond what we've seen until now. So what we want to do is build a blockchain in space and why until now we were thinking about space

as something that is really at the like state level or like very big enterprises level um innovation that's the classic like imaging location services with GPS weather forecasting communication all these are great but what if it's so easy to launch satellites that we can start thinking about completely new things. So the physical isolation in space gives us really unique advantage here because essentially what we see is that every node that exists on a satellite with the right configuration can serve as a tamperproof environment like a trusted execution environment aboard a satellite.

So that's kind of the question of like are we in that place now that we can start thinking about space as a real environment to iterate and try and build meaningful and high security guarantees things for crypto to really gain adoption. And that's what we're building with space computer essentially a distributed compute that expand the world computer concept into the space. And with that exactly on uh the architecture and how we're implementing I'll pass it on to Philip my co-founder.

Great. Thanks Daniel. Uh thank you for being here and uh as Daniel was mentioning the vision is very optimistic. We have a lot of work ahead of us and I will try to show how we actually plan to unroll it step by step to make sure that space computer can become a reality. Uh important to mention that uh during the great intro we are building on the shoulder of giants with respect to their achievements that enabled us to even think about uh achieving what we want to do and I will try to showcase on how we plan to combine all of the individual components together to make it a reality. So firstly as Daniel was mentioning what is actually space computer. So we plan to be a new robust uh layer one in space. Uh the motivation is is clear. Physical isolation try to

not only focus on the on earth use cases but we see a potential of future space use cases where individual satellite itself will be providing execution environment capabilities. Uh for those that are not familiar with the concept of TES, you can think of it like a confidential compute that can be run on a malicious envir in a malicious environment where either the host or the hypervisor can be malicious and each of the satellite itself can actually provide this space capabilities due to the onboard trust execution environment provided by the hardware but also have strengthened threat model due to the aspect of the physical isolation and of course in order to make it a whole and full-fledged uh infrastructure we cannot only rely on a single satellite, but we

need a network and constellation of satellites that will be providing the robust layer one infrastructure combining the individual components together. And of course, as a part of the use cases we envision is a full aspect of economy in space where we definitely start to see that more and more orchestration and settlement can be happening directly in space without the need of actually communicating with the on earth components as well. So where we currently stand as of now is that we have a single satellite that already enables us quite a few capabilities and in order to interact with the space infrastructure we are using uh our gateway called orbit port where orbit port itself uh is a infrastructure that we currently operate and is able to communicate with our infrastructure in space. Uh as of now uh

we envision that this is going to stay like that. We are aiming on making it more trust minimized as I will be highlighting on the other other slides and aspects of the orbit port will be extending to accommodate for future use cases when it comes to interaction with a with a aspect itself. As of now uh we support the first use case which is a generation of a cosmic torren number generator that can be already fetched and used in use cases on earth but in the future we want to expand it to other applications such as custodial wallet and more. And of course on the aspects of the user we want to provide a similar capabilities by interacting with infrastructure where the earth users should be able to interact and send the transactions either through uh orbit port or other type of uh components to

make sure that their interest in their transactions or others will be later on communicating with the with the respective hardware in space and at a similar time since we'll be relying on a different communication media either starink or many others we also think of it like a dedicated backbone or backup links that can be using different type of communications than we are used to from traditional internet aspects as well. Since we are envisioning to scale our system, we don't only want to provide uh the L1 capabilities that will be providing the the aspects of the robustness and the security for the settlement. But of course we need to scale through layer 2 solutions on earth since we are in a medium to short-term

definitely communication and computation constraints and for that we are uh getting a lot of inspiration from the layer 2 solutions from the ethereum ecosystem that can be deployed uh either in the optimistic or secure lab setting and later on interact with the L1 itself. But of course what we are also interested is to provide some form of novelty with respect to providing additional capabilities uh to the L2s on ref as well irrespective on which layer they are settling. So on that front uh we definitely see uh several use cases that can be relevant either in the new cryptographic primitive in the direction of space te and many others e providing communication and computation resilience to the users on earth itself.

Of course, the main assumption for this type of solution is that we will have a full-fledged uh constellation which is currently something we are actively working on and I'll be providing also at the end a short road map on what is what are the next steps for us moving forward. Diving a bit a bit more into the current use case and actually how the architecture itself look like with respect to orbitport what is happening under the hood. So uh we see ourselves mainly as a software company with respect to designing and providing the protocol that later on others can interact with and for that we are partnering with different type of infrastructure providers that can provide us the hardware but of course we also want to have the hardware specification and the requirements under our control and provide them and

communicate with individual partners that can later on launch the equipment for us. One of those providers is currently Optos orbital that is already providing us with a satellite we are communicating and using um which currently provides us the access to the randomness from space and I'm looking into the details on how does the interaction from the user perspective on earth looks like is that they are communicating with the gateway called orbit port to fetch the relevant information and provide it to the users on earth. As of now we are able to provide you with integrity meaning that you know that the data that you are com that you are receiving on the on earth are coming from the respective infrastructure but we plan to also expand it in the upcoming weeks uh by providing end to end confidentiality

meaning that you actually encry the data in the space and you as end user get these data on earth in a full confidential fashion in the end to end flow as well. So this is definitely something that uh we are actively working on and as a part of the end to end flow we are also actually looking into the applications and use cases of tees on earth that can provide us this trust minimized setting where actually orbit port itself can provide and be a trust minimized component with respect to the communication between the user on earth and the infrastructure in space.

As mentioned already, we are now looking into the aspect of randomness that can be used in several use cases where we are also looking into some aspects of innovation on how to integrate this randomness um into cryptographic or consensus protocols for example in the direction of peer selection uh consensus leader election or some cryptographic protocols for example third signing that are relying on randomness that could be actually uh getting the randomness uh true randomness especially important from space itself. Looking at the road map, orbit port is the the enabler of many of the use cases we envision. As mentioned, as of now, we have the aspect of the CTRNG, but we envision many many other use cases beyond either if you look at the secure

vault or space vaults. when you think of it uh that can provide custody uh by running your key generation process among multiple satellites where each of them is also having the T capabilities that can be also very interesting for aspects of DAO where we have certain guardians that can be definitely relevant of course for the future D5 applications if you want to have a location where you can store your collateral in a secure fashion but of course in general the trend is towards programmable cryptography so we are planning to provide a SDK tooling to users that can later on play and interact with the capabilities provided by by the L1 and the individual security services and the future uh looking is definitely towards a celestial compute marketplace where you can do and have more autonomy in space among individual

providers as well. With that in mind, uh we tried to outline the road map where uh the immediate next step after we have published our vision blue paper last year in November is to now really nail down the specification of our consensus and other components where as of now we know that we plan to be EVM compatible with respect to the smart contract capability platform aspects and we want to now finalize the aspects on how the network itself will look like with respect to the peer-to-peer layer consensus and how can we all combine these individual components together as which will be outlined in the white paper coming out in the upcoming months and moving forward we definitely see some interesting use cases either for

key management as I was mentioning in direction of custodial wallets but we definitely see key management being a problem in many other settings even in within the the ecosystem for proper key management services themselves uh since we see that already as of now we have an access to a single node we definitely want to expand on uh many proof of concept implementation that can be done uh on a single node itself. For example, by deploying a full-fledged uh client providing execution and consensus itself combined where we can actually already deploy the smart contracts and later on uh start to look into the aspects of proof generation for aspect of verifiability but also for the aspects of finding a way on what is the cost our infrastructure can handle and look

beyond also into the L2s Avs and similar for scaling on earth because the users on earth as a part of the journey should be interacting with the L2 components and later on uh get settled on the L1 itself and of course the ultimate goal is to provide a full-fledged L1 constellation where we will have multiple nodes multiple satellites in our case that will be each of them having full-fledged uh space T capabilities and will be able to to reach consensus on a confidential platform itself.

So we are a space computer. Uh we already have available uh for public demo uh the aspect on how can you interact with the CTRNG and uh if you're interested feel free to follow up and also catch us up. Uh Daniel Yango and others will be here throughout the whole week and we'll be definitely happy to answer some of your questions. Thank you very much. [Applause] I I just wonder how um how do you test that the TRNG in the satellite is still functional and how do you even know that a TRNG is broken or

is actually still giving you supposedly to randomness? So we are currently looking into the aspects on actually how to verify it. So there are like good test suits from NAT for example that you can actually use to verify the entropy of the random numbers you are getting and of course we have to make sure that we'll be doing this regular checkups because it can keep changing but the general assumption is due to the fact that the satellites are running in space you are having the cosmic radiation actually is affect affecting your random number generator. uh the the fundamental assumption is that because of that we are able to get the random number generator but definitely uh there is some more work to make sure that we can continuously verify that actually this is the case because maybe due to the radiation itself maybe some components

of the chip that is actually later on providing and fetching the noise itself can get compromised. So definitely it's it's a continuous process for now it's just the initial idea that if we have it what can be done with it. So it's not far it's a little bit farther from making it as a production ready solution but definitely as a part of exercise is definitely something that we see as a good potential on how to make it possible and the fundamental aspects are in place. We just have to now make sure that we walk the talk and put it into production. One last question. Yeah, I think it's um really cool project by the way. Thank thanks for coming. um actually came to the previous talk and stayed because uh it looked cool. um in uh look my my opinion of of

web 3 is that it doesn't matter how fancy we get with the blockchains or the D apps or anything the infrastructure screwed and it's all Amazon AWS all this stuff it's all centralized so it's cool that we're trying to take it off planet so my question is more around how diverse are satellite providers I know nothing about that so if we get up there in space is it all just Elon Musk owns the whole thing are five 6 10 100 different satellite providers if you could help us understand that. Yeah, it's a great question. Uh I think that's the vision. We see that as Daniel D was mentioning now it's a perfect timing because we see that fortunately unfortunately many of the aspects of space are very little for many governments. So they actually try to really push on the frontier on how can

we decentralize the front of having our own manufacturing capabilities having our own capabilities on actually providing communication. uh so that way of course we can support Starling we can support all these are operating in different jurisdiction as well and of course we see that uh other countries or continents and trying to provide different communication uh settings as well. So definitely on that part we see that due to the nature of the space race becoming now I would say very prominent due to the capabilities and the control over it that individual players want to have it's also something that enables us to actually have the centralized consideration. And so with respect to the infrastructure we want to build, we want to have this stuff as a part of the requirement meaning that uh we are able

to communicate with different providers. So that way we cannot just like get censored on a single one. But of course uh this is a the end end goal that we that we envision. We'll be taking individual steps to maybe at the beginning be more centralized on the front but definitely looking into the future and uh decentralizing a lot of individual components uh before we can go go live. It's it's becoming it's very competitive. So you have more and more providers like initially yeah like everyone's relying currently on Starlink for on SpaceX for launches but um Blue Origin is becoming operational. I think everyone is like in Europe people are concerned about depending too much on Elon so they're starting to encourage the industry here. China obviously so it's like it's growing. It's maturing

slowly but yeah it's like at the moment it's still not as diverse as you would imagine that that you'd want but yeah. Thank you so much for your presentation. Thank you very much. [Applause]