New Ethereum talks, every Monday. The week's conference uploads by event, in your inbox.

Loading player…

ThunderCore Consensus Explained by Samuel Harrison of ThunderCore

ETHGlobalFri, Oct 9, 2020, 12:00 AM

Samuel Harrison Director of Strategic Partnerships & Product Marketing at ThunderCore gives a talk at ETHBoston on the difference between traditional proof-of-work consensus and ThunderCore's consensus protocol. ETHBoston was an ethereum focused hackathon that took place Sep. 6-8, 2019 at Northwest Building, Harvard Campus. ETHBoston was a hackathon organized by ETHGlobal and a team of hackers, developers and enthusiasts interested in building and creating in the ethereum ecosystem. The ethereum focused hackathon took place for 3 days at Northwest Building, Harvard Campus in Cambridge, MA. For more information visit eth.boston and ethglobal.co

Transcript

[Music] have a technical presentation to walk there with you how you would employ it back so it's gonna be pretty straightforward it's just laying up a little difference between traditional work consensus and London were doing with our consensus protocol no code is in there it's all graphics and images so it's really something that you can see and I can walk Jane 101 creeper so if you guys want to go into the work and that could also give a little treatment but if you want this community I can also have to walk through let's do this wait no how many people we have here that are actually here versus the ones that were reported in the back we have like two so my name is sam Harris I am the head of product marketing that doesn't work thunder core is a smart contracts platform or second generation platforms some things that set us apart number one we start off with being completely IBM compatible so in the you be built on Imperium can be ported over another for very pretty the two other points are other ones are kind of interesting our transaction peoplein we're looking at 1200 transactions per second with a second combination which means that if you are in future using these things then you'll be able to have a very seamless experience for example just two weeks ago I was owned other than Seattle Comicon and I ran something called drive sale by our Tichenor people have applied died there and they're able to do it so quickly it was like a real-world experience with a ventilation cavity so the question is is how is chi that that quick in fastboot would that level using a consensus radiation so our consensus protocols called palette PLA it's the completion of Elena she's work that she did under Thunder Ella and this is the full implementation it's called power most of us here probably don't actually need to go through this just a really quick overview about you-know-who footwork first-generation block joins they are powered by a Google Earth inside this house very groundbreaking it's designed in such a way that any node but Lourdes would stand up watching the nodes in the network have to compete to solve hard cryptographic puzzles the first one to solve the problem gets to extend the watching and they earn a prize for their hard work but it's slow you know it's low because this data you can company throughout the entire yeah right and having so many notes compete to state so it's bad for you so so to solve these problems we created the pala consensus algorithm in pala we take these abilities specific proposers that have the privileges to extend the blocks themselves to make sure that the proposers are honest there are lots of us to learn on by committee of voters the primary proposer will rather than create a new block and then the sense that you block over high-speed network connections with all the different voters and these voters respond with notes and if two-thirds of the voters vote yes then the block is notarized and is added to the chain when we have two blocks on to change then that want becomes finalized meaning that it's not going to change ever and it'll be the beautiful part of the beautiful record now occasionally of course the big question is okay how do we trust that composer if that proposal crashes it is not a mine it's a delicious what happens is we move on magnets index propose our line where voters and then for two thirds of those votes are collected then the next one goes to becomes a primary composer and it is continues from where the last week very very so we're able to achieve this consistency which is solving the double spending problem meaning that only one joke essentially one person in the time and were able to achieve the Lycans because of somebody that something goes down then we have somebody on automatically to fall back on to continue and always progresses unlike proof-of-work we are through state protocol so the composers and the voters are selected using a pursue state election process where anybody become a voter over closer by staking in within Thunder jump in terms of the election period and in return they receive rewards like you would have noticed any other state state based on that behavior is you know double signing or being offline not being available to be punished by slashing those funds because safety so this is the way to which we were able to do watching biltema college it's fast it's secure and decentralized so I just wanted to run through really quickly kind of what that was if you have questions if you want to build something our portal is available the developers portal ocean lot more robust information bell ringers under quite calm and then you can follow us in discord which is where a lot of the technical questions happen the telegram Channel there's a lot of conversation about trading a lot of telegram in Skopje and then of course I'm ready actually not that's kind of us since that you guys any questions if you guys want me to talk about anymore that was just my quick presentation well what's your lazy what's the second six seconds for sure we use the finality in under center for so a lot of attacks whatever longest six seconds if it's one related longer than six seconds and that triggers a new composer most the time any other questions twelve o'clock regularly and our average is regulatory transactions per second I've seen a magnetized 1500 I don't know if that was [Music] everyone's happy that means big-time people has a twenty four hundred transactions per second throughput but then they have millions of millions upon millions of users yeah the best according to the DAF radar the number one used watching application in the world it's gone back and they have 2400 day later so that's stable right like where'd weak people keep on talk about scale people talk about scale but the reality is is that bigger problems not scale at caesars we could stick at the level we are right now and you can stick workers here in the transaction we just can't deal [Music] and I know that so I know that we have a bunch of research that's coming out in the next three weeks we have a white paper on Paulo we have several other things that are coming out of all these so the way to this work that we had we went our mainland with live in February using the Thunder La consensus mechanism which bifurcated are changing two things ace fast happen slow so that when the proposer would fail rather than switching to the next proposed online it would fall back to the Imperium chain until a new composer could be elected from the key the mechanism for switching wasn't defined over the course of the last six months we realized I was an elegant and frustrated so we finalized this proposal switch mechanism and so there's going to be a hard court in and over it is I have to double-check the date there that's gonna show that it's gonna be the hard for that up crazier consensus magazine team so as far as all the details on the fault tolerance or any other metrics you'll have to wait for my big girl now I know that you're in there I've seen them I've read them that work right injectable then this was a little bit faster than the 25 minutes I was alone

Automatic transcript — names and jargon may be misspelled.