Beyond the Chain

Scaling and sharding with Beniamin Mincu, co-founder of Elrond Network

September 17, 2020 clare saunders
Beyond the Chain
Scaling and sharding with Beniamin Mincu, co-founder of Elrond Network
Show Notes Transcript

In episode four of Beyond the Chain, Chris and Nash co-founder Ethan Fast speak with the Beniamin Mincu, the co-founder and CEO of Elrond, a distributed transactional computation protocol that relies on a sharded state architecture and a secure Proof of Stake consensus mechanism. They discuss scaling, sharding, user experience and more!

Chris: Welcome to Beyond the Chain the podcast from Nash that takes a deep look into blockchain technology and connects the dots between decentralized finance business and the wider world. I'm your host Chris Fenwick. And each week I'm joined by a Nash founder and the special guest this week here with me is Nash co-founder Ethan Fast.

Ethan: Hey Chris.

Chris: Our guest today is Benjamin Mincu an early blockchain pioneer in Europe founder and CEO of Metachain Capital an investment firm and also co-founder and CEO of Elrond, a distributed transactional computation platform which relies on a sharded state architecture in the secure proof of stake consensus mechanism.

Beniamin: Hello Chris.

Chris: So Ethan you are a technical advisor for Elrond. How did you become aware of their projects and get involved with the team?

Ethan: I've been honored to be an advisor to Elrond for a few years now. I actually first met Benjamin through Fabio. 

Beniamin: We met in San Francisco at a Neo event a long time ago. And from that point we kept the conversation going and we got into contact and had some discussions and it's great to have had you as an advisor .

Chris: Benjamin would you like to tell us a bit more about your background?

Beniamin: Yeah sure. When I was little I basically was one of the. First to have a computer I think in my class even though we were in Romania and so forth I have three brothers and my father I remember very well brought us a computer and it was super exciting for us.when you started playing of course breaking it very fast and then started repairing it and learning a lot of things. Then the internet came along in Romania. Interestingly enough we have one of the fastest internet speeds in the world. So the computer and the internet essentially opened a different world to me to,to lots of people here. And then from that point I remember that I always had a very deep passion for everything related to computer software that when I discovered deeper technologies and startups. I've become extremely fascinated especially by technologies that have the potential to change this scope of society. So this is how in 2012, 2013 I stumbled into Bitcoin. And this one was extremely intriguing because for the first time it combined not only technology but did combine technology with economics 2016 together with my brother I essentially started the fund through which we invested in probably 20 to 30 startups in the blockchain space especially yeah. To help them move up a lot faster solve some of the problems and also get a very deep overview on what was happening. Long story short after all this time in um at the end of 2017. Seeing that the most fundamental problems in this space we're not getting solved. This is where I decided that the best way to address some of these problems and improve what we thought was really necessary for the space to move forward was to just form a very strong team. A team who can literally build rockets and address some of these problems. So this is how we started Elrond.

Chris: Could you maybe tell us a little bit more about Elrond and particularly what the vision of the company is and what your aspirations are?

Beniamin: Sure. Our vision with Elrond is to build a global transparent non-inflationary financial system and then to give anyone anywhere easy access to this system. And we're trying to specifically do this first by building this most advanced state sharded proof of stake architecture.So it's a blockchain that can process transactions at one 1,000 fold. The performance of Bitcoin or Ethereum and at 100 or 1,000 X lower costs then Bitcoin and Ethereum right now such that at this point we can process more than 15,000 transactions per second with six second latency and 0.001. cent transaction costs. We're super excited about this. The most important part is that this network is already live. So it's not just theoretical discussion. secondly in addition to the network through which we want to bring broadband speed To the blockchain space. We also understand that there needs to be a sort of radical improvement in terms of user experience because if every user has to understand what we are doing here and work with the technology that is currently very very painful to to work with it will not have any kind of adoption. So we've put a lot of thought into this and created which is a super simple user interface that will give access to all the key features of Elrond's blockchain. And we'll open this access to anyone in the world with a mobile phone. So together these pieces should really accelerate the great things we've been seeing in this space. And we welcome any effort for the same. It's the same goal. So this is where I see a lot of commonality with Nash and where we're excited about how things are progressing there as well.

Ethan: [00:05:58] one thing that's always really resonated with me about what you're working on at Elrond is your focus on the user experience and really making it possible for many more people to interact with this technology which sort of to date still is kind of riddled with all these issues like needing to understand cryptography and how you should store your secrets and all of these things that for truly global broad adoption people are just. not going to understand right? They need to be solved. So I think that's like really a core focus both for Nash and for Elrond. um you know kind of with a different hat also Um you didn't mention this quite as much my feeling is you're actually bringing some of that to the developer experience as well. Um so I haven't personally spent a lot of time developing anything on Elrond yet. Although I hope that changes pretty soon. but from what I can tell it looks like the tooling that you're building out there is really impressive too. So you know kind of the users as developers is also a nice theme. I think that that you have going.

Beniamin: [00:06:56] For sure for sure. And thanks a lot for bringing that up. I do believe that our team has put an enormous effort into solvin. like I said the victimology problem first and making sure that there is very high performance that you can work with that makes sense not only in processing capacity but then also in economics of the transaction costs and so forth but then moving one step further we've seen a lot of effort put also in the design. Of how the developers will interact with the architecture. And this is not only in the tooling also in the economics design. So for instance we have this really cool feature of providing 30% royalties that are built in for anything. The smart contract that is built on Elrond So the author has a building of economics model there too to make sense of a business that he wants to build. And whenever he has adoption this adoption is immediately and directly reflected in money that he will be able to use and then I think too that we indeed have Wasm VM that we're building on so that people can build stuff and make sure that they can do it in C++ and in Rust.

Ethan: [00:08:20] Yeah for sure. You know one very kind of special thing that Nash does is the security of the accounts through an API key setup using threshold signatures to allow people to interact with the exchange without taking the risk that whatever bot they're using has access to private keys that they can then compromise the funds. And so to build this stuff out we had to go through a pretty complex protocol and implement this for Neo Ethereum and for Bitcoin. whereas on Elrond basically the scheme you're using allows threshold signatures kind of as a fault right? So if you take a private key and then you do the simple transform on it you get these two keys which can then produce signatures which then can be combined. And there's very little work involved. It's the developer experience around that is super nice. That's the kind of better cryptographic design that makes building these other technologies on top of it so much easier. So we're looking forward to integrating with that because that will be a much less pain than it was for these other blockchains.

Beniamin: [00:09:22] Yeah for sure. I get a sort of boost any time when we discuss with someone that can also look under the hood and understand some of these technologies that have been built. It's fairly rare that people can see the nuances of these technologies. And as you rightly point out. When you look at the specifics and nuances that you discover how many cool things you can really build and how much easier it is especially if you've gone through the same process with Bitcoin Neo Ethereum and then you also have a reference point to compare. There's a lot of use cases that can be built with this new primitives that we have and we already see a good number of developers trying things out and and going through them but it's especially interesting to see someone with your background see these nuances in a different way so I'm very much appreciate your feedback there.

Ethan: Yeah for sure.

Chris: So Beniamin you talked about how Elrond is trying to solve scaling problems with Bitcoin and Ethereum. So a lot of people now will have noticed that Ethereum fees have become very very high because of all the activity on the network due to DeFi . Maybe we could step back a little bit and you could tell us a bit more about what it is that makes Bitcoin and Ethereum slow and inefficient. And also talk a little bit more about the things that Elrond is doing to try to solve these problems.

Beniamin: Sure. So the idea is that Bitcoin came along and offered a way of transferring value without a trusted third party. This is the contribution that the blockchain brings. it can sort of be seen as a database that In contrast to the normal database it's operated by many self-interested arbitrary individuals. What matters is that there's an economics design on top of the architecture that keeps a lot of people. incentivize to go towards a particular direction. But the problem with Bitcoin and Ethereum especially is at the base layer they're very constrained in terms of processing capacity such that the Bitcoin can process something like seven transactions per second and Ethereum can process something like 50 transactions per second. And the important point there is that the design on the network only allows for all the participants in the network to process the same amount and quantity of work. So there is no parallelization in this architecture whereas what we've done with Elrond looking at this problems from many different angles and we discovered that there was something much deeper there that essentially if you can parallelize transaction processing instead of processing everything in a serial manner with every participant processing the same amount of work you get to a very different setup where you can process two or three orders of magnitude more transactions to begin with. So to simplify sharding anyone who's looked at databases knows sharding is a way of partitioning the database so you can parallelize processing. In blockchains there's three types of sharding. It's a state sharding network sharding and transaction sharding state sharding has been typically the hardest problem to solve because it's in a distributed setting. You have many complex problems that you need to address and from what we know Elrond is the first architecture to have solved this and have a working model in production that has both cross shard execution for transactions. So coming back to the scaling problem. Elrond addresses this through state sharding where we're partitioning the network and having a transaction prioritization so that we can do the work a lot more efficiently and with a lot less cost. The one last thing that I would add is the adaptivity what we have is adaptive state sharding which means that this design allows us to start with only two shards and process 10,000 transactions per second but then it also allows us to change the number of shards dynamically real thought such that either there is demand we can immediately update the network as it is. So yeah. I'm curious if this clarifies the questions and um if Ethan wants to add something to that.

Ethan: [00:14:26] Yeah sure. Maybe let me just restate it really quickly to make sure I understand. And help other people understand it too. So with sharding you can almost think of it like you had one blockchain everybody had to agree on the state of the world and instead now we're going to split the state of the world into multiple different pieces. And as long as you're only interacting on one piece, that piece can decide the state of the world itself at whatever speed it can. And so we don't have to wait on all the other States of the world because it only cares about what's going on within itself. And then you just have a coordination issue where you need to get those shards to agree on some common state. Which I think is like the Metta chain. And so having those different shards um basically you can make as many as you want theoretically. as long as they don't have to interact with one another you can basically make it as fast as the number of shards you have. Like just add more shards everything gets much faster. what you mentioned that was super interesting is this idea of the state sharding. So to try to unpack that a little bit I would phrase that as, you know people are pretty familiar with like smart contracts. On Ethereum for example. so I have just like an ERC 20 token right? So this is like any token on Ethereum. I want to move it around basically that token maintains like a yeah. Database or an index of who has what? And so when I moved this token from one person to another. What actually happens is like the smart contract managing the token says okay I'm going to adjust this user's balance to this and this other user's balance to this right.It's actually more complicated than that but that's that's kind of how it's working. And so if we're talking about sharding if that token lives on one shard then as long as you're moving it around on that shard that shard can not even really have to interact with other Charlotte's right.

Cause all the state lives there. it can do whatever it needs to do. And then the Metta chain can keep going. So what happens when I want to use a piece of that state like on another shard. So for example I want to send that token let's say to an address on another shard does that make sense as a problem? How would you describe Elrond's solution to that?

Beniamin: [00:16:26] Oh yeah I do agree. It very much makes sense. So the idea is for our architecture you basically have this metal chain. that essentially keeps a track of everything that happens in the. Shots but it keeps dragging of Vanya of very very small cool amount of information the smallest necessary amount of information that it needs to store in order to keep the sharks in sync.

Ethan: [00:16:59] Some event is triggered (sort of you could call it an event) which then because it touched another shard it's labeled such that it needs to be processed by the Metta chain. And so then the meta chain processes that it validates it as part of its consensus. And then once that happens the other chart is now aware of this event and can sort of understand the state based upon that. Is that roughly what's going on?

Beniamin: [00:17:22] yeah. So the the idea is that you basically have the sender shard processing the information And then um at that point the meta chain just make sure that what has been processed is not in any way objected to.

Ethan: [00:17:38] I think there's another big component to the scaling solution. Elrond has that. I didn't hear us talk about it a little bit which is the proof of stake component. That's also an important piece right?

Beniamin: [00:17:48] For sure. So we have two very important breakthroughs that we've built this entire architecture on. The first one is adaptive state sharding. And with this we essentially scale things significantly to parallelize transaction processing. And then secure professors is this mechanism through which makes sure that we process the transaction very efficient, fast and then very secure.So we have this random sampling of the consensus group so that you have 500 nodes in each shard then from those 500 nodes you have 61 that are randomly sampled every five seconds. You never know who will be the consensus group. And then even if you would want to collude and do try to put someone down acts on the network you only have those six seconds.And after that you're again randomly reshuffling the consensus mechanism and then at the end of each equal which currently is defined as a day. You have again one-third of knowns in each shard being reshuffled to all those shots just to make sure you have another layer of security there. So we're whereas the state sharding adaptive state charting model ease well actually responsible for scaling the secure proof of statement mechanism is responsible for making sure that what needs to be processed can be in a very secure of very fast and very efficient manner.

Chris: [00:19:22] so we've been talking about scaling from a tactical perspective but maybe we can talk about scaling now in the business sense. So have you faced any barriers so far in running our run that have required you to pivot or change your design and what are your plans for scaling and growing Elrond? The run team for our own platform in the future?

Beniamin: [00:19:44] So we have a very special target that we've set so it works as a clear guiding direction. This is that we're trying to bring the next billion people in the blockchain space. And with such a clear idea in front of us then yeah a lot of the immediate decisions either immediate or medium long term become. It helps us prioritize some of the things. So we're of course at the point where we have launched the main net um and we have this really cool new application that is coming out and it's being tested which is uh Meyer and what's that we are trying to do is approach things in two ways on the business side. First we have several startups and projects that are trying to build on Elrond and use the smart contract design architecture and royalties that we've put in place. And we're quite excited about that. And on the other hand we have this very very clear path. Towards a lot of users adopting block chain via Meyer and minor needs to diversification.

That will likely be the very simple user interface that people discover crypto through it should be the simplest way they interact with. And we're now in a process of. Rapid iteration where the first version of the first Meyer alphabet version has been out. We've been going through two cycles with a bunch of great people from the community that are super passionate closed wrong and have been very early on the waiting list. And beyond that we're trying to launch the first version as soon as possible and start growing very very fast. There's a very high demand for the application. So after two days of launching or reviewing a preview of the application we gathered 50,000 people on a waiting list that wanted to get access to Meyer.So we're obviously quite excited about this and have wanted to push things so we have the first version as soon as possible and then focus all our energy on growth on that direction. The team has been growing and I think we are one of the smallest teams in the space that's doing a fairly large number of things. We have a lot of really really resourceful people that are putting a lot of energy. but we are about to grow the team very fast. I think it's always remarkable how fast the teams grow but then the output or the productivity does not necessarily grow which the team's growth rather diminishes immediately after you grow. And then only after some time you can grow the output as well. So at this point we were really focused on speed. And then once the first version of a Meyer is out we're going to expand a lot more. Ethan I’m curious how things look on your end. Have you guys expanded a lot? Do things move really fast on on your end as well?

Ethan: [00:23:03] Yeah. Yeah. I mean growing the Nash team has definitely been a really interesting and exciting experience for us too. I think Elrond is is working in traditional more physical kind of office space - correct me if I'm wrong about that. But Nash, it has actually always been a remote company. This is something we talk about a lot. And there's some really interesting challenges building remote companies. I think you know all of us as we grew the team to like 30 it was pushing 40 people the coordination issues when you're dealing with asynchronous remote teams are just really really challenging. And we had to design new structures and processes around that. So that's something we've been learning a lot from over time. I think the world has kinda been thrown into a remote sort of context in a way right. With the pandemic so maybe you guys have also had to deal with that a little bit but for us that's been the kind of really interesting challenge there.

Beniamin: [00:23:53] I do agree that this is super challenging especially when you compare it with uh in house set up where you meet the people and so forth. And it's been a very interesting experience for us because we have an office and there's a very good energy in the office where people meet each other. They discuss things problems get solved and I've always felt that. whenever I left the office there was this very fast streamed river that's flowing and you're stepping away whenever you're in the office everyone is going there's a lot of things happening and there's a great energy there but then when you leave it's a strange feeling for a few days and then you get used to it. But interestingly enough when we had to switch to a remote set up with the pandemic things have worked extremely well. I would say we've put in place a few things like a daily sink to begin with. and then free flowing communications can join a Hangouts group and sit around and have a discussion and so forth. And it's remarkable how much smaller things have helped that keeping touch, keep seeing and then work through any other problems. And then interestingly enough we had an even higher output uh productivity and energy during the time where when we were remote. 

Ethan: Yeah. Yeah. I think we've also kind of had some uh productivity boost from the pandemic. I think there's definitely an effect where when there's like literally nothing else that you can do people actually are more inclined to spend time on their computer which is I guess nice for a startup.

 Chris: COVID does actually highlight some interesting societal changes and the shift towards qremote work. But perhaps we could talk a bit more about the broader vision for Nash and Elrond. So what kind of concrete societal changes do you potentially see coming about as a result of blockchain technology?

Beniamin: The current version of the economic system that has some problems that I think cannot be solved from the inside it's so broken and the feedback loops are not there to allow it to move in the right direction. So to this effect we believe that we can come and compliment. The current economic model we have with the digital model. Whereas I've said in the beginning we can have a global transparent and non-inflationary financial system to which people can connect, interact with and so forth. The second point is the distribution problem and we specifically tried to tackle this through Meyer because it was through Meyer as long as you have internet and the mobile phone you'll be able to access this new way economic system. then we also have the. Eagle which is the native currency of Elrond which will give people access to a different store of value. One that is more robust. More sustainable and whose design and rules are always transparent and communicated and discussed with the community and where improvements can always happen. So this is the immediate stuff that we are looking at. Tens of millions and hundreds of millions joining this space just because there's an obvious utility and incentive there. And hopefully during the next three to five years our goal is to make sure we accelerate and onboard maybe 1 billion people to this space by making the technology totally invisible. I'm curious Ethan, how Nash is trying to push things on this front.

Ethan: [00:27:45] Yeah. Yeah. I think there's a lot of shared goals there. I think in particular just making the technology more accessible. It is really the key thing that we keep drilling on. Empowering anyone anywhere to be able to access this sort of new class of digital assets and sort of all the things they allow you to do.

that's what we're trying to do. And um I think the main thing that Al Rhonda or that you guys are doing differently is that right? in addition to focusing on the interfaces around these technologies you're also focusing on the core blockchain infrastructure right?

The core infrastructure that allows all this to happen in the first place. And so by solving those scalability issues there. that's obviously very helpful so maybe we can already allow someone to transact with cryptocurrencies just with what they would think of as like a typical user account but not going to solve the problem of you know their transactions taking like several hours with Bitcoin or you know these kinds of slower feedback loops and weird error States. I think there's a strong kind of complimentary vision with our two companies.

 Beniamin: Yeah totally agree. Totally agree.

Chris: Benjamin is there perhaps anything else you'd like to discuss or talk about?

Beniamin: I'd say that the most important thing for Elrond right now is that the network is live. We'd been working super hard to bring this to the market. Now our core focus is on adoption. Adoption through Meyer will come with the speed that is impossible to be brought via other means. So this whole vision of improving the financial ecosystem is exactly what we need during the next period. And it seems to me that Elrond is at the cutting edge of just laying the necessary pieces so that not only Elrond but anyone else who has as a particular vision. through which we can accelerate things can build things on our own and improve the lives of users and people anywhere in the world. We welcome people in the community to build smart contracts on Elrond and start deploying cool applications there. They can also become validators and there's an entire universe there as well. we're super excited about all of this, one of which is the integration with Nash as well. And just having eagled dare to be tradable will be really cool for our communities. So we are quite excited about all of this.

Ethan: Yeah we're super excited too. And I look forward to developing more on Elrond.

Beniamin: Awesome.

Chris: Yep. Thanks a lot. It certainly sounds like there is a lot of shared vision between the companies thanks a lot to Ethan for joining us today.

Ethan: Thanks Chris. it was great to be here. I'm glad we were able to have Beniamin on. 

Chris: Yeah. And thanks also to Beniamin.

Beniamin:Thank you very much Chris and thanks to Ethan. Lovely to have this conversation.

Chris: And thank you to everyone who listened to beyond the chain subscribe to their favorite listening stream and you won't miss an episode. You can find the full text of today's episode at