ThunderCore Basics

hi thanks for coming so my name is Jim G I'm walking thunder from the coins in here today I will give a talk about the Sun called our launch and so I assume some people already cover us and even played with us and for this talk I just want to give people some brief introduction about technology and I understand how what are we building and how we build it how it works and give people some details about 7 o'clock here so very quick recap what's the what we are building so Sonico is a general purpose public chair and it's had very high throughput right now are we already launched it on man that it achieves a 1280 PS and if you already played with us do some of the transactions you can see the confirmation about 1/2 seconds for one transaction and it's fully compatible with EVM earning EVM a serum smart contract the apps can migrate to our chair in minutes it's scalable its support to high throughput without failure and the sex for our co-founder professor Lane she just a citizen here there is all the work has been mathematically proved there to secure so that's what sonic highways ok so maybe people we asked why sonic also fast and how does we achieve decentralization from the sonic all so I hope after this briefing talk you can find his answers I hope you can understand how it works before we dive into details about how summerhall works we can do a quick recap of some of the traditional proof of walk block teamworks so basically we know there for example for a serum their huge number of nose distribute all of the wart and they're connected by a p2p network if you want to submit a transaction play a game or send the tokens to a friend you suffer me to one transaction to one of the nodes in the network and given the most of the honestly host knows agree on the log we can achieve the consistency and the transaction can incorporate Sassoon and we achieve the loudness then eventually we got this chain was built up so however because it's a bill p2p Network it has very limited bandwidth even some protocols some protocols like a nakamura owner it end cannot fully leverage it in network bandwidth such an bandwidth issue and introduced a long longer latency for our transactions and we can see the how slow the throughput of the network and how long of latency and we may rethink about a massive remodel of this block to network and to improve the scalability and rethink about incentives and economics based on maybe POS so here is our answer some the real a paradigm that's a protocol suite named after Stan Arella actually the Thunder is also the prefix of the prefix of Thunder that's how our company's name come from Thunder call or sundered token something like that okay in our network we also have some full note just like a serum we have accelerator and a one or more than one accelerating our network and multiple from tens to hundreds committees of validators that's architecture of Thunder Allah or something called network if you want to send a message or your own phone to submit a transaction definitely you still need a senator to our phone out then the first disturb we call the propose the accelerator will receive all the transactions which are related by the phone out it's just to insert this insert of this transaction in one block buta block and assign it you the PRS key and the recorded is proposed and the standard is the first block it has not been confirmed confirmed send it to all the validators and all the validators will validate all the committee members will validate the transactions to see if the legitimate or valid if everything in file they will sign this propose to block individually using their PRS key and ascended back to the accelerator this step we called a vote suppose accelerator received more than 3/4 of them a vote we the the model think this block contains all the validate transactions and we will generate to the finalized a block and we call this this step is the notarization and thus the blocks will be inserted on the channel so you can see is this a single round of the single round of the vote and is that's a basic way that and and it's a Harper spoke and network architecture rather than just rather than a p2p network that's a basically how we achieve the fast or high throughput and achieve 1200 the TPS in our men at but and actually this is already conservative number in our test environment where it's easy to achieve 2,000 or even more but to guarantee reliability in our minute we just guaranteed 280 TR 200 sorry 1200 TPS form a net that's how basically how we generate how we generated blocks how we builded the chair so maybe next question is okay who dominate the whole process how do we guarantee the consensus how we guarantee the consensus the consistency so I want to notify one thing here accelerator here he does not vote he is just a coordinate coordinate the committees has a power to vote or to make the consensus on the blocks so basically the first layer of the decentralization of our system relies on the committee's as a mention that we can have tens or scale up to a hundred of the committee's and that they could be everybody everybody who holds a standard token can be a committee member through the bidding and the rewarding process suppose you hold us on the token you can send a bidding message to our fools say hey I wanted to be a committee and here but you need a staking your tokens I want to to ten thousand or 101 million tokens to be a committee and we will run a bidding there's a whole network will run the bidding protocol based on the tokens you have staking or some other variables to determine who could it be the committee for the next round after each committee member they vote on the each block they will earn the credits based on the transaction fees so more votes and the most taking more tokens in use taking the more rewards you were receiving that's a basic idea about how to how anybody can be a committee member and to get rewarding from the whole system okay then that's that's how the thing works maybe they're necro veterans how okay this this block chanced on the collection achieved 1200 TPS we called it a faster pass then the next question is how do we know the masterpath walking properly and we have basically two mechanism to guarantee everything in fine first one is a live message a line message you actually use it send it by the accelerator so this is just a snapshot for series of locks on all faster pass it's a hash over serious blocks and will be posted on another ground another trend we call the slow chain which is well known to everybody looks like a plan to choose so so far we choose a serum as a as a slow channel so if the fast path goes well if accelerator is not crashed if the mode if the committee's more than three-quarters the committee is still walking everybody should have seen a live message on the slow chair periodically if the along the city is missed that means something's wrong with our fast pass in a meanwhile if you submit your transaction what if the accelerator all the committee members never vote on your transactions or something censorship happened or something wrong so if your client you can send a year message also to the slow chair that's the year message actually is the incra in a raptor transaction you intended to submit it to our Fast Pass you just say hey why do you guys do not to process my transaction here's my transaction you do not show up on the first apprised of us right now I suffer me to the solution and that's a serum you cannot manipulate everybody should see it so the Year message which will be inventively put on the solution will be money to the everybody including foo note included committee so after the year messages are the slow chair it should be processed sometime later if other knows including committee we assume most of them honest they cannot see the year message or the transactions are processed they will stop work they will think something's wrong with accelerator odd oops sorry okay so they will see something wrong with a Fastpass and doing some censorship or the dishonesty of things if either one of this thing happened we caught a fall back through to happen so most of transactions almost behavior will fall back to just solution so we need to do some another recovery walk to either find a new committee or find a new accelerator to recover the Fastpass that's how we handle decentralization first of all everybody could a committee member through the bidding process secondly if you're worried about the accelerator we first know we we could have mountable and we have this fallback recovery mechanism to guarantee the same accelerator or the whole network can do things something wrong because we are waiting for the line message on the slow chain to guarantee the network is working to guarantee the lammles and also we have the year message to guarantee every your single transaction should it be considered equally with others if this one should it be processed so that's how we achieve decentralization based up using the fallback recovery mechanism between the Fast Pass is a slow chair okay quick recap for santarilla in two sentence first of all we have a single round of voting that's a way that's with the way we achieve high throughput once things are good and once these go back we will use a slow chat for for back again in this case is a serum that's a grown to choose that's a back-up plan for Thunder Thunder call as well santarilla however however the reality is not is that amazing fallback is always a cost and we do some we did some experiment the accelerator switch takes about 10 minutes and in some worst case if even slower think about that all the recovery should happen on the slow chair and the urine is low chair is really slow if I'm the user of the Sun call if the fallback ever happens I will see a long latency of everything because nothing good has happened that's something really painful for our users so back to October last year October Elaine and our research team have a new development of the Sun derailleur something real a paradigm then we have another protocol called pili and a Paula especially Paula so that's the new Fast Pass consensus so basic idea that previously maybe on the first pass there's only one accelerator but actually we can support multiple accelerator on the first pass and we can do edge our rotation or election on the first pass for the accelerator as well so think about a lot previously we only elected committees and we did not to talk about accelerator we think is neutral we only have something to monitor the accelerator when we did not talk who could be a accelerator but Impala everybody can be not only a committee also could we have a chance to be accelerator and you could be a backup for the next accelerator and we do the accelerate accelerate election on the fast paths as well so that means if for back something happened something's wrong with accelerator happen we do not have to fall back to slow chair we can do the recovery on the fast bus as long as a committees most of the committee's still survive that to save a lot of time lot of the latency for all users and for our system and and we we have something more enhancement recap if there's only one accelerator only one accelerator can propose the blocks that means we only have one pipe if we have a multiple we actually can do something in pipeline that not only improves the lowliness improve the latency for the fallback recovery also could it be opportunity to improve the throughput of the whole system as a whole network that's we called Paula and we already have open source so then another question if we have Paulo do we still need slow-jam do we still need a slow chair because you think about that if you need a slow chance you need a take care of two system that makes implementation really complicated and solution is a back-up plan honestly we do not want touch it because that means something wrong happened and that means a bad user experience so another thing for is another benefit of part of that it can help us to do the decentralized monitor on the fast path solely just like accelerator rotation and also we we may still need it but all needed absolutely only as to the very extraordinary circumstances for example the entire committee corrupted that means the elects elevator election cannot happen or some basic functions cannot performed on the first pass that's the only worser case so basically we may still need a solution but the power makes things simpler we stay on the fast path for most of time and the most operation including the accelerator rotation will happen as fast pass only and also improve the throughput of the whole system okay above that's just basically how santarilla sunder call works with a standard call only the Sun derailleur and the sounder call with a Paula so we already did a lot that we already have a lot of something I would announce here first of all open source plan so we already open source Paula consensus POC this Wednesday I remember yes so everybody can access our Paula POC code from the github and we are approaching to the sounder Kofu node folder open source in q2 2019 hopefully by the end of June if not in July so we are testing actually I was looking at a bug fixing reports yesterday lots of funny thing happened for fun ordered stuff and besides that after we open source fou note we will have a plan to welcome everybody to roam the fou note stand a call phone out on your machine and I heard from the product manager so at the beginning we will at least the support 100 external node all over the world and we will release more information and the tell people the exact a program for that and after open thoughtful note we also go ahead an open source committee an accelerator code and we will see we will see our migration from sander Ella to Paulo in our open source program you will see our peers or panel opens polyp you see staff will eventually migrated to our product our code in hopefully by the end of the year this year that's informal on official announcement sorry also we are calling for external committee as I mentioned everybody could it be a committee definitely if you have more tokens you have high probability to be a committee for those guys who has less committee who has less token if but you still want to benefit from the external parameter rewarding we have another program I will cavallo omit just after this slice so we already planned a external committee program and if you if you pay attention to our plant or media we already reserved seats for different people 25 reserved seats for nonprofit organizations may be investors of may be in Velenje universities 25 seats for staking pool service those fours are retail investors or the people with a small amount of tokens we also reserve 50 50 percent of the seats for open bidding for everybody you can just try to it and we want to this incentive maximum to encourage honesty participants and the high performance participant participation one half of ones or high performance think about that our accelerator waiting committee to vote latency computation all those stuff is if the votes come very late it will just downgrade our throughput so we actually looking for honest and high performance participants in the meantime so that's some challenge for us let's see what we can achieve okay stick input service if you have tokens you can earn more tokens through this program no matter we will run some committee seats on top of this staking pool service you may not occupy a single seat but you may share the city with some other people and you can earn a transaction fee and you can also we also set up a 100 meaning t-tom the token for the rewards and at most if you staking out tokens for more than for one year you can get to the 30% APY from from this pool and I think this one will be launched very soon in couple days Oh as I heard well we are already down or all the testing it's just in the internal beta tester right now I hopefully we can you can see it maybe a monk on Monday we also are working on the cross chain asset so we know a serum is a very very fantastical economic ecosystem programming languages applications and its own ecosystem sums of non fungible tokens and the stable cones and a lot of digital assets also survive together with a serum we want to support we want to leverage and support this ecosystem so as I mentioned in the second slice sander core is a serum compatible things day one and the technology with superior to the to the serum and we want to give a serum community better use experience since we are more scalable and prank and provide a high throughput we want to support all the digital sets answer is serum and give the users developers opportunity to import or export access to sonic all like either like yeah see turning tokens and non-visible tokens and are the staple cones so this one is already on the engineering plan so we will soon we will see the details very soon okay this may be good for developers and defense of the Thunder we already launched the Ambassador Program Ambassador Program is the program we tried to attract as a technical talents and the join us as a community so here will be stands for Thunder to the local community all over the wall are all all over the world so you can see the details from this blog link the primary link here and we already got to some applications and we already have 80 University ambassadors will join us maybe students fans or even professors and we want to praise awareness of the standard call to campus and promote thunder call through this Ambassador Program ok again so thank you everybody for coming so here's a basic timeline so about the company so we have the u.s. office opened on February 2018 and in July we have the Taiwan office but actually also to our last year we launched our test net and then in the q1 2009 team we launched our mandate so it's already there for almost a three months and you can see it's it's grow very fast we already have huge amounts of the Tau and the some stuff i sorry i do number ml is that pin numbers 11,000 tau okay and we we are welcome everybody to come to play with us deploy out the apps and we will have workshop maybe today right we have another workshop arrived last ok sorry so and we will also another workshop like a trifle in this year and welcome everybody to come with us to come to play with us and one more thing I want to emphasize we are purely compatible with a serum any is serum compatible program or the app can migrate a to standard calling in minutes all the Asura existed in developer tools like a travel like a like remix can integrate a wizard chair without almost result any change you just need to change your RPC URL director to our Santa Claus stuff then it's done so that's something we are proud of we we give the most convenience to our developers and our users beside that you will see more program of for the developers we try to bring users to developers and help to build it a bridge between the users and it develop not only the cryptocurrency geeks offense but also regular users actually just a couple days ago our CEO Chris Chris won't give a talk and consensus we right now we are specifically talking to the game gamer players regular game players not only cryptocurrency a creep equipped or game players we want to bring more and more people onto our chain as well as to the blockchain community and bring more opportunity to developers to enlarge their user space of their applications okay that's all my talk we still have two minutes for question go ahead an accelerator from censoring some of the validators that go through it can possibly change the outcome of a boat like it eats they won't it looks like on the diagram that the validators are routed through the accelerator so committee yes okay so first one first one because accelerator plays a key role here although it does not vote but it's a collector transaction in a proposed blocks it needs very largely resource to do it if it's a it's a very weak one definitely will Dunkley the process so the benefits or the incentive for the accelerator bidders or the participants that it can take more percentage of the transaction fees from the block that's the basic way we how we incentivize people to participate in as accelerator he has another question okay the second one so you mean the censorship right the censorship or why not okay so if the censorship of the traditional of the regular transactions we have Yael message we have your message if your transaction never seen being processed you put your year message on the slow chain so everybody can see a year message including the committee the committee also money to the Yale message on the slow chair then it wait it will way to Fox accelerator if accelerate is still honest or no censorship he may missus miss this transaction for whatever reasons but it also need a money to the slow chair if it detects that's your message around the slow chair and to will pick it up and the check if it's already process or not if not it will put it in the next block immediately and the comedians should have waited for these transactions for the next to propose the block if let's say for long time the here message you still not to processed that means a warning so accelerated tween something wrong maybe do some censorship it's a committee if the committee realized that it can stop working okay I do not agree this trust to this accelerates something wrong they stop avoiding is out to means a live message will not be generated because the live message is a snapshot of a serious blocks okay thank you