2018 May 07 - Volume.2 Issue.18

Goto Full Issue

2018 May 07 - Volume.2 Issue.18

Goto Full Issue

Radix, Hedera Hashgraph Eye Main Chain Scaling Solutions

Radix and Hedera Hashgraph, both of whom are building blockchain networks in an effort to address the scaling woes the industry has so far seen, spoke to Diar on the progress of their platforms. Radix demonstrated in a live test that it can sustain more than 20,000 transactions per second. And Digital Currency Group backed Hedera Hashgraph have also partially tested theirs. Both networks have taken different technology and governance approaches are slated to go live in 4Q18.

Bitcoin’s vision for fast peer-to-peer transactions with low processing fees has not come without its fair share of problems with its popularity growth in 2017. At price and transactional volume peak in December 2017, fees reached more than $30. Since then businesses with mass volume have optimized their transaction management by implementing batching and Segwit support, lowering the median fees to less than $1 again.

And Ethereum also struggled similarly with the popularity of CryptoKitties, the first well received blockchain game, that boomed at the end of last year. As the game consumed around 15% of the whole network traffic, median transaction fees increased to upwards of $3 – 20x current fees. While the bottlenecks haven’t been resolved, development teams are looking to address the issue with Bitcoin’s Lightning Network and Ethereum’s Raiden Network – both off-chain solutions.

The opportunity to create a scalable Blockchain network from the ground up addressing scaling directly on the main ledger hasn’t been missed however. Diar spoke to the CEOs of two upcoming public networks that are being built specifically to address scalability of traditional blockchains - Radix and Digital Currency Group backed Hedera Hashgraph.

Piers Ridyard, CEO of London-based Radix tells Diar that there will always be a bottleneck on the Blockchain and while “it’s of course possible to scale by sidechains or state channels or other solutions, there will always be a bottleneck on the main chain” – options that the two majors have had to resort to. The comment comes from the company’s own experience who have been in Research & Development for the better part of 6 years.

Radix initially started with looking at how to scale Bitcoin in 2012 but then it “really quickly realized that the blockchain architecture itself wouldn’t work.” Radix since then has set its sight on building a protocol with an ambitious mission that would “give anyone anywhere free access to digital economy”, easy application deployment through APIs, and would also be interoperable with current technological and economic infrastructure, such as Point-of-Sale.

Radix also looked into using a Directed Acyclic Graph (DAG), an alternative architecture used by IOTA and Hashgraph, but found that “it doesn’t shard in a way that it’s possible to always guarantee that a double spend will be spotted unless there is centralized control.” The company, instead of meshing together an existing architecture with an existing consensus mechanism, started researching and building new architecture with the goal of supporting mass use of the network simultaneously.

The Radix team found that the only possible way to achieve this was if the architecture sharded by default and the consensus mechanism worked across the shards by default as well. The platform, which would separate a large database into smaller manageable parts, has now been live tested, and was able to achieve more than 20,000 transactions per second (tps) in comparison to Ethereum’s 25tps (see table).

“Even if you were to store as much data as Google, it would give you about 2MB of data per one shard so we still wouldn’t be at the level where any of those shards is too big for even very low powered devices to use. The true decentralized network is when any device, even with very little resources will be able to serve as a full node on the network” Mr Ridyard says.

Apart from scalability, Radix has two other objectives in order to build a global decentralized protocol - developer friendliness and user friendliness. Mr Ridyard says that writing secure smart contracts with no extensive experience doesn’t exist at the moment, which is why problems have been seen in Parity and with the DAO. With the massive overlap in the code of smart contracts “a lot of developers are writing the same code from the scratch, which makes it very difficult to build something secure”. Radix opted to have built-in base level functions in the protocol itself that can be called directly via the APIs, which will “lift a massive burden off developers in building on top of the network”.


In order to make the network friendly for average non-technical users, Radix is in talks with manufacturers that build plastic cards with a secure element, which is capable of generating signatures for transactions without revealing the private key. Mr Ridyard says that “it is exactly the same to how any hardware wallet works. What we have done is instead of requiring people to buy an expensive hardware wallet, we made the signature scheme, that is already well established for doing cryptographically secure signatures, compatible with the Radix ledger. The starting point is to make the ledger compatible with the hardware that already exists, which will help a lot with mainstream implementation.”


Radix is indeed aiming to give developers tools to build decentralized applications and to deploy tokens. The network is scheduled to go live in Q4 of 2018 along with the utility token issuance whose purpose will be to “reward people for building and deploying on the network rather than buying in and speculating.” A difficult set goal, in today’s environment at least, is to keep the volatility of the token low - a fact that the company recognizes. “We don’t think that we can create a token that will have a predictable value into the future until the entire protocol has real utility to the society.”

Hashgraph, developed by Dallas-based Swirlds, was initially only deployed in private, permissioned-based networks. In March, Swirlds announced that Hedera Hashgraph is spinning off to create a public network, in which Swirlds will only play a role of the licensor of the technology.

Radix Formula For Adoption

Global Scalability
• Demonstrated the global scaling ability

Developer Friendly
• Built-in smart contract functions called directly via the APIs

User Friendly
Ledger that's compatible with existing hardware
Secure cards that mimic hardware wallet grade security
• Decentralized exchange (DEX) built in the Radix protocol

Hedera Hashgraph Model

Performance - Successfully demonstrate the scaling ability

Security - As long as fewer than a third of coins are held by bad actors then you can’t attack the network.

Stability - By having patents on Hashgraph, the platform is guaranteed to not fork into a competing platform and provide transparency by release the source code for public review

Governance - Council with equal voting rights consisted of the world's biggest companies

CULedger, the global consortium of credit unions, announced that it will use Hedera for cross-border payments. Mr Harmon tells Diar that the reason why CULedger chose Hashgraph is because the performance is “orders of magnitude better than all the other networks.” He added that “compared to other permissioned networks, Hashgraph is not susceptible to denial of service attacks. Hyperledger and Corda have a leader and you can do run a DDoS attack against the leader and bring down the entire network.”

Diar spoke to Mance Harmon, CEO and co-founder of both Swirlds and Hedera Hashgraph and confirmed that it was always the intention to build a public ledger “but bootstrapping a proof of stake (PoS) system required that we have a Council of global bluechip organizations involved in the public ledger”. Mr Harmon says that Hashgraph wanted to build some reason for the governing members to want to participate, which meant that it needed to start with enterprise use cases on the private side of the market.


 The Council would give voting rights to 39 organizations “that would not likely collude with one another”. Hedera is eying the largest globally distributed corporations across 18 industries. The company has commitments for more than half the governing body, but does not have the full 39 as of yet, which it aims to announce in full by this fall. Apart from Swirlds who is a permanent member, the remaining 38 organizations can only serve a three-year term with a limit of two consecutive terms. As long as not fewer than a third of the tokens are held by bad actors, the network can’t be attacked, which is why the Council will hold two thirds of the tokens until such time that the value of the token is high enough and broadly distributed enough that it is not possible for bad actors to get a third of the tokens.

Mr Harmon says that the “Council creates the most decentralized governance model of any other public network on the market today. In all the other public networks, the decisions are made by either the core developers, miners or a foundation with only a few members.” All the decisions will require the 39 members of the Council to agree. "There are no master nodes. All nodes are the same. There are no Coordinators, Leaders, Masternodes. The nodes are all on par with each other" Mr Harmon confirms.

Mr Ridyard does err on the side of caution on the breadth of the decentralization of Hashgraph however. “Swirlds and other 38 entities will absolutely control the network and to me that doesn’t seem like something that should be a starting point in building a platform that we want everyone in the world to use. To me, it sounds like you are putting an incredible amount of power into the hands of very few people. As for the technology, “Hashraph ultimately has a really strong centralization control and there is a really good reason for it. Sharding a DAG without centralization control means that there could be really easy double spends.”

Hedera tests that it can reach anywhere from 50,000tps and up to almost 500,000 tps. One problem, however, is that the transaction speed estimates do not include time to process transactions. Mr Harmon tells Diar that “Hashgraph as an algorithm has fantastic performance in posting transactions in order but the bottleneck comes in with the signature verification.”

Hedera did work with one of the global graphics processing unit (GPU) manufacturers and wrote a library that does signature verification on the GPU, which ends up eliminating the bottleneck. Another aspect of processing a transaction is how long it takes to update the balances in the list of accounts. Mr Harmon says “our experiments internally suggest that it’s really minor in terms of CPU and storage requirements compared to everything else. We really don’t think that that’s going to be a bottleneck. We are expecting to achieve hundreds of thousands of transactions per second.”

Mr Harmon says that there probably won’t be a general retail Initial Coin offering (ICO). “We are domiciled in the United States, which of course puts a lot of constraints on us. What we are doing is demonstrating over a period of time that the tokens are actually being used to create the economy”. Hedera plans to go live in Q4 2018 and release a third of the tokens in the first five years. When the remaining two thirds of the tokens are continuously released “because there is no longer a threat, the funds will flow into treasury and we could end up having billions of dollars in treasury that can’t be touched.” When the need arises, there is a natural buyer of the cryptocurrency to help regulate the volatility and supply.

wdt_ID Platform Ethereum IOTA Radix Hedera Hashgraph
1 Consensus Decentralized Not Fully Decentralized (Coordinator Nodes For Now) Decentralized Decentralized
2 Governance Anyone Can Propose Eip, After Consencus Is Reached, Ethereum Foundation Implements The Change Iota Foundation (For Now) Radix Team (Eventually Decentralized Governance) Council Consisted Of 39 Who All Have Equal Voting Rights And A Limited Serving Time
3 Code Open Source Open Source Will Be Open Sourced Source Code Will Be Released For Public Review
4 Inception 2013 2015 2012 2016 Private, 2018 Public
5 Architecture Blockchain DAG Tempo (aBFT, aBFD, Sharding) DAG (aBFT,Sharding)
6 Consensus Mechanism Pow (Pos Planned) Weight Is Added To Attached Ancestor Transactions Tempo Gossip About Gossip And Virtual Voting (Hashgraph)
7 Max Transactions 25 Tps 800 Tps (Will Increase With Time) 20,000 Tps (Will Increase With More Nodes) 50,000 Tps Up To Almost 500,000 Tps (For Achieving Consensus On Transaction Order And Timestamps, Estimate Does Not Include Time To Process Transactions)
8 Security 51% Of Bad Actors More Than 33% 51%-99% Of Bad Actors 33% Of Bad Actors (Eliminated By Not Releasing More Than 33% Of Tokens)
9 Patent No No Patent Pending Patented
10 Business Model - - - Cut From Each Transaction Fee

Receive Diar Every Monday – The Digital Assets & Regulation Trade Publication

Something went wrong. Please check your entries and try again.


Disclaimer: Unless otherwise specified, the content of the articles published on www.diar.co constitutes intellectual property of Diar Ltd and may not be reproduced or republished in whole or in part without prior written consent. The information contained in the articles published on www.diar.co does not in any way constitute financial or investor advice and is only intended for informative purposes. Readers may not rely on such information to decide on investment or financing options or otherwise rely on such information in making decisions with monetary or financial effects. Diar Ltd does not accept any liability of any kind with regards to the validity of the information or with regards to any damage suffered as a result of reliance on such information. © 2018 Diar Ltd. Contact: newsdesk@diar.co