Hashgraph vs EOS

by May 22, 2019Hashgraph vs

Hashgraph: Hashgraph is a proof-of-stake consensus algorithm with DAG-based distributed ledger technology, that uses 39 term-limited elected Governing Council members with top speed, global scale and bank-grade security.

Hedera, the public network built on Hashgraph, can scale to millions of nodes and uses a utility token, the hbar, for network and transaction fees as well as security. Hashgraph offers a cryptocurrency, smart contracts, and distributed secured file storage in a multi-use platform.

EOS: EOS is a delegated proof-of-stake (DPoS) consensus algorithm with a blockchain-based distributed ledger, that uses 21 Block Producers that are elected for their role by the community with modestly improved speed from Bitcoin and Ethereum, and with probabilistic security with BFT quality of finality.

EOS is one of the top 5 cryptocurrency networks that offers a digital coin as well as a Smart Contract platform for games, decentralized apps and decentralized services. EOS was able to raise over $4 Billion USD in an ICO at debut and has a thriving developer community rivaling the Ethereum community of developers.


EOS Price chart provided by Trading View

Coin vs Coin



100,000 TPS/shard
Parallel Processing
Low Latency (seconds)
High Throughput


~3,996 TPS/shard
Serial Processing
Modest Latency (3 min)
Low-Modest Throughput
15 Block Confirmations
(0.5 second blocks)



100’s of Millions TPS/sharding
3 Phases of Node Topology
Massive Global Scalability
Minimal CPU Requirements
Minimal Bandwidth


10-100K TPS/sharding
21 Block Producers
Multichain Scaling
High Bandwidth
Limited by small node #



Math-Proven aBFT
DDos Resistant
Sybil Resistant
Botnet Resistant
Firewall Partition Resistance
Low Collusion Risk of Gov Coun


DDoS Vulnerable
Sybil Resistant
Botnet Resistant
High Collusion Risk of BP’s



Governing Council
39 Members w/ 2.56% vote/each
Max of 2 elected 3yr terms
Anti-Fork Controls

ECAF Governance
21 BP’s w/ 4.76% vote/each
Less Nodes>>Higher Risk
Of Collusion



Fairness of Order
Fairness of Access
Fairness of Timestamp
Fairness of Node Entry (Phase 3)
Mathematically-Proven Fair


Low Fairness of Order
Modest Fairness of Access
Low Fairness of Timestamp
Low Fairness of Node Entry
High Risk of Censorship
No Formal Math Proof



Proxy Staking Earnings
Virtual Voting Algorithm

Delegated Proof-of-Stake
Bonding Stake Earnings
(3 days)
Gossip Protocol



High Efficiency
Lowest Energy Consensus


Modest Efficiency
Modest Energy Consumption

Regulatory Compliance


SEC-Compliant Fundraise
Not a Security per Hinman
No ICO or ERC-20 Token raise
Gov Council removes illegal files


Non-SEC Compliant
ICO Fundraise
ERC-20 Token raise

Development Team


Hedera Hashgraph, LLC
Very Experienced Team
Small Developer Group
Growing Social Media


Dan Larimer, et al
Block.One Company Founders
Based in Hong Kong
Large Developer Group
Cayman Islands



No Central Authority
Open-Review Code
Patented Algorithm
39 Nodes Phase 1
1000 Nodes Phase 2
1 Million+ Nodes Phase 3


Low Decentralization
Open-Source Code
21 Block Producers
21 Full Nodes
Light Nodes
(not for consensus)

Use Cases


Smart Contracts
Secured File Storage
Financial Services
Certification Services
Government Services
Healthcare Services
Commercial Services
Social Media Services
Many more


Smart Contracts
Decentralized File Storage
Games & Gambling Dapps
Social Media Services
User Identification



ACID Compliant
Formal Coq Proof
FLP Theorem Compliant


No Formal Proof



Hashgraph Speed: The Hedera Hashgraph public DLT is capable in a single shard of performing at least 100,000 transactions per second in a globally distributed mesh of peer-to-peer nodes. This is the current top speed of any public DLT to date. The throughput is high due to parallel, asynchronous processing and has low latency even at high scale, measured in seconds.

Hashgraph Speed Limit: The speed of the network is only limited by the bandwidth and speed of one’s internet connection. If internet connection speed increases, the speed of hashgraph will also increase in speed proportionally.

hashgraph speed


Transactions occur asynchronously, meaning that maximal throughput can occur without fixed time intervals or bundled transactions. A global 5G network with low earth orbit satellites could allow anyone in the world to host a node and earn hbar from staking.

EOS Speed: The EOS blockchain is a Delegated Proof-of-Stake platform that has claimed the capability of a maximum 3,996 TPS in a Permissioned-Network of 21 Block Producers. Due to the probabilistic security of blockchain, nearly 15 confirmation blocks are currently recommended, producing ~3-7 minutes latency with each transaction.

EOS Speed Limit: Due to the underlying serial processing of blockchain DLT, a chain of fixed, synchronized transactions must occur in order to achieve consensus. This does not permit asynchronous transactions like the Visa Network or Hedera Hashgraph.

Therefore, EOS will always be limited by its own chain time of 0.5 seconds, with multiple confirmations and bandwidth-dependent design. A real-time network monitor is available, with a current average of 50 TPS.

Link: https://eosnetworkmonitor.io/#

Hashgraph Scale: Hashgraph is able to perform at 100,000 TPS in a single shard (database).  Hedera the public DLT built on Hashgraph has the ability with multiple shards (database divisions) to expand to 100’s of millions TPS with marginally decreased performance and latency.

Hedera is able through PoS consensus to scale to millions of nodes and allow any one the ability to earn hbar in their wallet with minimal CPU requirements, memory and bandwidth. These features of a public cryptocurrency are fully able to accommodate micropayments and major global enterprise solutions.

REF: https://www.hedera.com/hh-whitepaper-v1.5-190219.pdf

EOS Scale:  Since EOS is a public blockchain DLT that uses 21 permissioned nodes, or Block Producers, scaling will become an issue as the network grows. Consensus is determined by this small number of nodes in order to increase speed, at the loss of decentralization.

Moreover, in order to scale, sharding solutions would also have to determine a probability of finality of transactions, which would take multiple confirmation blocks (~15 blocks for ~99% security). Thus, in order to prevent a double spend, EOS has to vote on the state of the blockchain then send it to the shard for approved use. This induces latency with each shard.

The multi-chain structure of EOS allows for less network congestion compared to Ethereum. However, since consensus is limited to only 21 nodes, network congestion is conceivable at global scale, while decentralization is low.  

Hashgraph Security: Using a proof-of-stake model, Hedera Hashgraph has achieved bank-grade, asynchronous Byzantine Fault Tolerance, which has been formally-verified in a Coq proof. Karl Crary, Associate Professor of Computer Science at Carnegie Mellon University verified the aBFT property mathematically in a computer proof. The entire network by extension is also aBFT, maintaining on-graph consensus between shards at scale through the use of state proofs.
hashgraph security
This level of security is the highest possible and was not thought possible for a public network since it is 100% final within seconds. Speed and scale are completely preserved without reducing throughput or latency. Hashgraph is immune to the trilemma.


EOS Security: Using a delegated proof-of-stake model, 21 nodes (Block Producers-BPs) are elected every 2 minutes, allowing a random node in the group to generate a block of transactions. This step takes 0.5 seconds, but must then be verified by at least 14 other BPs before there is a 99% assuredness that the transaction will become irreversible in the EOS blockchain.

This level of security is Byzantine Fault Tolerant (BFT) and is less than 100% finality. In order to provide high security, EOS has to produce more blocks (15 at least per transaction), which reduces speed to about 3-7 minutes for higher security transactions. For example, to send a large value of EOS ($1M USD) in one transaction may require 21 confirmation blocks, thereby increasing latency for a higher probability of security.


Hashgraph Stability: Hashgraph is managed by both technical and governance controls in order to prevent the network from being copied or forked (split into to rival networks). By technical control, no node could use the Hashgraph algorithm due to the immutable genesis address book and its reference to the founding network as primary.

Additionally, Hedera is stabilized by a 39-member, term-limited Governing Council comprised of multinational, multisectoral mega-cap corporations that host the Phase 1 nodes. Each member holds a vote with a 2.56% weight per member, serves an elected 3-year term with election allowing only two total terms.

The Hashgraph algorithm of consensus is patent protected, and has an irrevocable license in perpetuity to the Hedera public DLT. This legal control, under execution of the Governing Council is meant to prevent network forks. These controls prevent network splitting, which institutional research has determined as a major barrier in the enterprise adoption of distributed ledger technology. However, Hashgraph is open-review in order to maintain transparency of code, allowing developers and node hosts to ensure that the source code is safe and free of malicious intent.

Stability of the Hedera network is also resistant by several design elements to Sybil attacks (34% attack), Malicious Network Partitions, Round-Robin attacks (DDoS that follows the IP address leader) DDoS attacks, botnet attacks (multiple attacks from a single coordinated attacker through cloud technology).  Additionally, the network is immune to mining collusion and hardware arms races due to PoS consensus.

REF: https://www.hedera.com/council

EOS Stability: EOS is an open-source network, which allows anyone to copy and create a similar competitive network with minimal effort. This ability to create competitive chains leads some to believe that enterprise adoption will hesitate to build on public networks without firm controls on platform fragmentation.

Governance in EOS is voter driven by token holders, who vote on proposals that the EOS Core Arbitration Forum (ECAF) decides on. The Block Producers enact the decisions of ECAF but also have the ability to alter consensus with majority control. This in fact occurred to the EOS network, whereby the 21 Block Producers voted to block nearly 20 accounts. The community reactions was not favorable due to the censorship that was easily achieved by the small amount of Governance controls in the hands of few members.


Hashgraph Fairness: The Hedera public network allows fairness in three separate ways: fairness of access, order and timestamps. Fairness of access is based on the randomization of transaction requests to nodes so that no single node can block a user, without having another node available for it.

Fairness of order means that the Hashgraph algorithm used for consensus, arranges all transactions fairly into the data structure so that no user can bribe any node with additional fees into advancing to an earlier state. Fairness of timestamps means that any transaction sent by a user will be arranged based on its median timestamp of all nodes, in order to fairly assess the global time that an event occurred.
hashgraph fairness

These features of fairness of the Hedera public DLT will be mission-critical for applications in real time trade, auctions, stock exchanges, cryptocurrency exchanges and many more. Censorship-resistance is ensured by design of the Governing Council who are only able to delete illegal material from file storage for regulatory-compliance. The GC are not able to delete users, block wallets, reverse transactions, amend a smart contract or even revoke a user.

EOS Fairness: Due to the low number of total nodes allowed in the network, there is a hosting node bias to only 21 Block Producers who share in the highly profitable EOS rewards. Fairness of timestamps is low in EOS due to the nature of blockchain design.

EOS’s blockchain allows random order of transactions in a block without individually accurate timestamps. A certain amount of transactions will occur every 0.5 seconds, all with the same timestamp. Fairness of ordering is also not present in EOS due to the blockchain data structure.

Hashgraph Consensus: Hashgraph uses a proof-of-stake, virtual voting algorithm of consensus that achieves universal agreement on shared data by staking hbar cryptocurrency to nodes. A node processes hbar transactions by signing digital signatures (public keys) after a uses submits a transaction by signing their private key digital signature first.

Hashgraph Efficiency: Hashgraph has one of the lowest energy requirements of any public distributed ledger due to the PoS model, simple hardware necessities, and extremely small amount of work and data load per transaction. A typical hbar transfer uses approximately 140 bytes per transaction with an upper limit of 4 kilobytes for any transaction (i.e./ smart contracts, file storage).

EOS Consensus: EOS uses Delegated Proof-of-Stake, with 21 Block Producers that arrive at consensus to a high degree of probabilistic certainty after 15 block confirmations. Users vote on the Delegates and the top 21 winners become Block Producers. The Delegates that do not win receive a small award of EOS tokens. The Block Producers are rewarded in EOS tokens for signing transactions after one is randomly chosen to begin the next block.

The vulnerability with this kind of leader-based system, is that a DDoS attack could be aimed at the sole BP that is responsible for forming consensus. If this node’s IP address is discovered than the next assigned IP address could also be discovered, which could allow a ‘Round-Robin’ DDoS attack that could disrupt consensus.

Efficiency: EOS consensus is a more energy-efficient process than PoW systems, but sacrifices security by reducing node numbers to 21 and having a permission-based system.

Hashgraph Regulatory Compliance: The Hashgraph network has technical features that permit users and developers to bind both pseudonymous Hedera accounts to Verified accounts for know-your-customer (KYC) and anti-money laundering (AML) regulations. This is an opt-in feature and can be used by developers depending on pertinent regulatory requirements.

The cryptocurrency utility token of Hedera, the hbar, is not by any community standard similar to a security since it is used to protect the network. It meets the Hinman Token Standard Definition by being sufficiently decentralized and is not at risk for SEC securities fundraising since it used a regulatory-compliant SAFT with accredited investors. No ICO or ERC-20 token fund raise was used in Hashgraph’s founding.

EOS Regulatory Compliance: EOS is predominantly based in Hong Kong and ran the most-grossing ICO of all time with nearly $4 billion USD. Many in the community saw this as an excessive amount of early fundraising without having even a mainnet, using ERC-20 tokens for the ICO. This most likely would qualify as a security according to the U.S. SEC and would be at jeopardy of regulatory action.

Hashgraph Development Team: Inventor of the Hashgraph algorithm and Co-Founder of Hedera, Dr. Leemon Baird is a Professor of Computer Science at the US Air Force, trained at Carnegie Mellon University for his PhD, and has been a senior scientist in several private ventures. He has experience with previous successful new startup companies, holds multiple patents and publications in computer science, machine learning and advanced mathematics.

The CEO is a seasoned entrepreneur in multinational corporations, government agencies and technology startups. He is a former Founder and CEO of two additional prior successful companies acquired by private equity. His most remarkable achievements may be his roles as Program Manager of the US Missile Defense Agency, Course Director for Cybersecurity at the US Air Force and in a private laboratory in Machine Learning.

The team of Hedera Hashgraph, LLC is well-experienced, with master-level education with a proven track record. There remarkable funding in early phases is also a major achievement for the team. Many members have ostensible and visual profiles in LinkedIn and other social media outlets.
dr leemon baird
Since Hedera has not opened main net access for public development, the developers are difficult to estimate. However, to date over 300 separate dapp projects have been developing on the Hashgraph network.


EOS Development Team: Dan Larimer is the CTO and Co-Founder of EOS. He is well experienced in the cryptocurrency ecosystem having begun other projects such as Steemit and Bitshares. Many of the Block.one team behind EOS are experienced computer science professionals, investors and entrepreneurs. The community of developers is very strong with a large following in dapp development rivalling Ethereum.

Hashgraph Decentralization: Hashgraph is a distributed ledger technology with decentralized governance, nodal topology of 39 members (Phase 1) and can scale to millions of nodes. It is censorship-free, has pseudonymous accounts and has no majority-controlling agent. The risk of member collusion relevant to all other cryptocurrencies due to multinational and multisectoral distribution of members is very low.

Hedera Hashgraph, LLC is a network solutions software company formed by the Governing Council members and has an irrevocable license to use the patented Hashgraph algorithm in perpetuity. Due to the proof-of-stake demands of bootstrapping, the Hashgraph Treasury will have the majority of hbars for the voting algorithm. This has low initial distribution and will need clarification of future disbursement after the initial 5 year lockup.

It has been stated that the Treasury will not allow profiting by the Council, but will be used for hackathons, community development, software updates and node support. The Hashgraph Governing Council will only be stewards of the Treasury and will have to act in unison to properly spread hbars in the community to prevent Sybil attacks as the network grows and the hbar earns value.

REF: https://www.bloomberg.com/profiles/companies/1644627D:US-hedera-hashgraph-llc

EOS Decentralization: Due to the DPoS consensus mechanism with a low number of nodes in a permissioned-based system, is at risk of Block Producer-collusion since only 15 of the 21 would have to be compromise for an attack. Since the launch of mainnet, a unanimous decision was made by the 21 Block Producers to censor a certain of accounts for supposed nefarious activity.

This demonstrated the power of joint collusion in the EOS network and how it can be used to censor users on the platform. Overall, EOS has a low degree of decentralization by virtue of the trilemma that all blockchains suffer from. Speed is increased in EOS by reducing Decentralization.

Hashgraph Use Cases: Due to the multiplatform design with hbar cryptocurrency, smart contract functionality in Solidity, and secured distributed file storage, the Hashgraph network is capable of disrupting all business sectors including public and private industries. Being fundamentally an aBFT-certified network, Hedera Hashgraph will be able to accomodate global megacap commerce through micropayments, financial technology uses in digital asset and stock exchanges as well as public auctions.

Hashgraph is fast, fair and secure and has the highest throughput and best latency for a distributed ledger. This should allow global adoption in fields such as massive multiplayer online games as well as digital assets, allowing real time payments and transfers at the speed of the internet. The use cases are unlimited in fields like medicine and law, where secured and confirmed deletion of files, licensing certificate issuance and revocation with zero knowledge proofs would be invaluable.

EOS Use Cases: EOS is a leader in dapp development due to a thriving developer network that have built on the fastest of Smart Contract blockchains to date. User Identification, distributed file storage and cryptocurrency use cases are all possible on EOS. Many of the use cases that are possible on Ethereum are also similarly possible on EOS.

However, due to low speed and scalability issues, EOS may have difficulty scaling to massive global demand such as in the the credit card and payments industries. Also, EOS does not natively support fair timestamping and therefore would be excluded from public securities exchanges, auctions or real-time online games.

Ħello Future.