Scalability has always been a big question for people in the world of blockchain. What happens as this technology becomes more and more mainstream, with networks rapidly filling up with millions of users? A technology project that started out as a niche project between enthusiasts, can successfully scale to a global level? It’s a big question, maybe not so far, but that’s quickly changing with technology.
The Blockchain Scalability Issue
Cryptocurrency is becoming more and more mainstream. In fact, let’s check out the most popular blockchain network and how the network has gotten over time. This is a graph of the number of daily Ethereum transactions tracked year to date.
This may look very impressive, but here is the thing. The initial design of cryptocurrency was not meant for widespread use and mass adaptation. While it was manageable when the number of transactions was less, as they have gotten more popular a host of issues have come up. To compete with more mainstream systems like VISA or PayPal, the blockchain networks need to seriously step up their game theory when it comes to transaction times.
As an announcement from the Ethereum Foundation, the consensus mechanism of Ethereum network will shift from proof-of-work to proof-of-stake, as the event called The Merge. (The author suggests the audience read through the article on Ethereum 2.0 that is linked first) The merge will facilitate the whole system in case of scaling architecture. Ethereum has come up with a lot of scaling solutions which have either already been or are going to be implemented, fews like proof of stake, off-chain state channels, plasma, block size increase, segwit and sharding.
The biggest problem that Ethereum is facing is the speed of transaction verification. Each and every full node in the network has to download and save the entire blockchain, even new nodes still have to sync and keep up everything data on the network. And for Ethereum roadmap, “sharding should ship sometime in 2023 depending on how quickly work progresses after The Merge”.
The scalability trilemma says that as a blockchain tries to achieve scalability, decentralization and security, it will only be able to attain any two of these. With security being an essential requirement, there will be a trade-off between scalability and decentralization. Sharding provides a way to achieve both scalability and decentralization while maintaining security.
New networks which provide the exact same or similar smart contract functionality as Ethereum have been developed to fill the demand left by Ethereum. A sharding EVM-compatible smart contract platform, so far is not a strange concept anymore, we can see the precedents before like Ziliqa, Harmony, Near Protocol, Elrond (not EVM). Bitcoin and Ethereum compromise on speed (or throughput) and Layer 2 chains may compromise on decentralization. But since anyone can run a nod, the scalability trilemma is addressed, security is by default, and neither decentralization nor scalability are compromised. In this article, we will go through some basics of sharding definition and features, and the next gen Layer 1 sharding blockchain network - Shardeum.
What “Shardeum” Really Means?
Dubbed as Shardeum, the blockchain technology poses significant competition for its counterpart in the industry, Ethereum applied Sharding into. It could be said that Shardeum is the world’s first layer 1 blockchain that solves the scalability trilemma. Shardeum is an EVM, smart contract platform which allows linear scale using dynamic state sharding with low gas fee forever and aims to onboard billions of daily users and numerous dApps to Web3 in the future.
What sets apart Shardeum is aimed at performing consensus on each transaction while the other blockchains group transactions into blocks and perform consensus on those blocks. Shardeum does consensus on each transaction separately. This allows a transaction that affects multiple shards to be processed simultaneously by these shards rather than consecutively as with block level consensus.
Sharding is a common concept in computer science that is the process of splitting a database horizontally to spread the workload. Sharding will split up the burden of handling the large amount of data needed to proceed over the entire network and continue to reduce network congestion and increase transactions per second. What sharding does is that it breaks down a transaction into shards and spreads it among the network. The nodes work on individual shards side-by-side. This in turn decreases the overall time taken.
Sharding is compartmentalizing sub-networks, also called shards, where the nodes keep track of transactions in a ledger. A single node in a shard can approve, save messages without going through all the messages. Sharding is dividing a blockchain into smaller parts. So security is the main concern in sharding and how each node will communicate during transactions. Currently, it is quite obvious that sharding is able to successfully break down a particular network into more than just one. Each shard carries out the transaction seamlessly. This paves the way for a robust network that gets supplemented by all the computers that join the network.
Dynamic State Sharding
Unlike static state sharding (i.e Near or Elrond) where all the nodes in a shard cover the same address range, dynamic state sharding requires each node to hold a different address range, but there is significant overlap between the addresses covered by nodes. Although dynamic sharding is more complex to implement than static sharding, it allows for true linear scaling. Each node added to the Shardeum network immediately helps to increase the TPS, whereas with static sharding the number of nodes that must join has to be at least the number of nodes defined as the minimum shard size before another shard can be created. Only when another shard is created does the network TPS increase in a stepwise way with static sharding.
The transactions are committed first before they are grouped into blocks or partitions. And this enables the network to state shard and evenly distribute compute workload, storage, and bandwidth across all nodes on the network. Each node thus, will be able to process a number of transactions per second, and a new term is coined then, “transactions per second per node”.
New Consensus Mechanism
Shardeum uses a combination of Proof of Quorum (PoQ) and Proof of Stake (PoS) consensus mechanisms. Proof of Quorum means to generate a receipt showing that a majority of the consensus group has voted for the transaction. Each node in the consensus group signs the transaction hash and gossips it to other nodes in the consensus group. Nodes collect these votes, and when the number of votes is more than 50%, these votes form a receipt that can prove consensus on the transaction.
Shardeum will make use of both PoQ and PoS consensus algorithms to keep the network safe where validating nodes can be expected to stake a minimum amount of network coin to ensure they operate as per network protocols. While most platforms group transactions into blocks and apply consensus at the block level, Shardeum does consensus on each transaction separately. This allows a transaction that affects multiple shards to be processed simultaneously by these shards rather than consecutively as with block level consensus. Every node, essentially, will know that every other node in the group knows a particular transaction on the network. This enables a trustless collection of votes (or quorum) in the form of receipts. And when there are more than 50% of the receipts, transactions are confirmed. Thereby, finality is reached very quickly with low latency which helps to avoid network congestion.
This not only reduces the time to process the transaction even if it affects multiple shards, but also ensures atomic processing. Transaction level consensus eliminates the complexities needed to ensure atomic processing which otherwise are needed by block level consensus platforms. Thus, PoS and PoQ will render the Shardeum highly scalable and secure while fee on the network remains low and stable.
PoQ Compliments Dynamic State Sharding
Shardeum’s consensus is designed to increase tps linearly with each node joining the network. In typical blockchain networks, the state of the network does not change until a majority of the nodes agree on the current state of the chain. Thus, the computing power of unsharded networks is as good as the node with the lowest processing power. State is considered the latest record of information of the account balance and smart contracts on a network at any given time. So this is significant as Shardeum will be a dynamically state sharded network which allows to maintain atomic composability through transaction-level consensus. On Shardeum, transactions are processed simultaneously across multiple shards and nodes.
- Dynamic Sharding : The network won’t have a static group of nodes as fixed shards. Nodes in the Shardeum network are free to move around and accommodate more data as dynamic shards.
- Linear Scaling : Every node added to the network will increase the transaction throughput right away. It doesn’t have to wait for a fixed number of nodes to be added to the network for a new shard to be created.
- Autoscaling : Shardeum is designed to auto-scale. The network maintains consensus on how many nodes it needs based on the traffic. It can allow more nodes to be added to the network based on bandwidth required and shrink down during trough periods. This adds to the high level of decentralization that is embedded in the network already.
Take an example for this, assume Shardeum can initially process 100 tps and there are 100 nodes in total. That means the network can proces 1 transaction per second per node. When 9,900 more nodes join the network, it will be able to scale up to 10,000 tps. And this is where state sharding and fast finality steps in, 1 million tps could be breached in some cases due to the linear scaling.
Theoretically, Shardeum has provided the public network with a solution to scale up - sharding, but in practice that only happens when the network is real when there is enough demand.
Three Types Of Nodes
Shardeum is expected to have 3 kinds of nodes:
- Validator Nodes: These nodes validate the transactions in the network by participating in the consensus. They will have to stake SHM to be able to participate, and Shardeum will reward them SHM for participating (network rewards will come from predefined SHM emissions and the transaction fees earned). Validator nodes don't store the whole history, so they are lightweight.
- Archive Nodes: Archive nodes maintain the entire Transaction history. Archive nodes may or may not have to stake SHM, but they will earn a portion of the network reward to incentivize these nodes to store historical data.
- Standby Nodes: These are Validator nodes standing by in the network and not currently participating in consensus. Standby nodes help scale the Shardeum network quicker in times of traffic spikes. They may or may not need to stake SHM, but they will earn some of the network rewards.
But, how is the Shardeum network safe from someone taking over 51% of nodes? According to the research of the author, a cycle rotation, a random deterministic lottery to accept nodes into the network, an attacker would have to take over 51% of all nodes in the list waiting to join a network and the usual 51% of nodes that would be in a network. Nodes waiting to join a network are not part of the consensus. To ever have a majority of nodes in the network, an attacker would also need to have >51% of nodes waiting to join for a sustained period. Suppose you had less than 51% of joining nodes. In that case, statistics will be against the attacker. The network would accept most of the honest nodes over time. If there are 10k nodes in the network but 100k waiting to join, the attacker would need to maintain >50k nodes for an extended time until eventually >5k of them are rotated into the network as the network slowly cycles in and out its entire set of nodes. It basically solves the blockchain problem of security and decentralization.
The entire project is monitored and administered by a reliable Swiss based non-profit organization. The founding members of Shardeum are Nischal Shetty and Omar Syed. Shardeum team members also known as committers are distributed across the US, India and Myanmar.
- Nischal Shetty - is the Founder and CEO of WazirX - the India’s largest crypto exchange with over 10 million users. Nischal is a highly regarded entrepreneur with over a decade of experience building and scaling global products out of India. A software engineer by education, Nischal has also founded Crowdfire, a social media management web service with over 20 million users in the past. Nischal’s previous successes have landed him on the Forbes ’30 under 30′ list. A passionate blockchain evangelist, Nischal has been active in the Indian crypto space since its inception. Nischal’s mission has been to make crypto accessible to every Indian; he’s also been advocating positive crypto regulation in India with his Twitter campaign #IndiaWantsCrypto for over 1000 days.
- Omar Syed - is a blockchain architect who organized the Shardus project in 2017 to build a linearly scalable blockchain. Over the past three decades, Omar has been involved with helping large organizations such as NASA, Yahoo, and Zynga build scalable, fault-tolerant, distributed systems. Omar holds a B.S. and M.S. from Case Western Reserve University with specialization in Artificial Intelligence. Omar was involved with several start-ups including the first matrimonial website and the first stock sentiment analysis website. Omar along with his son Aamir invented the strategy board game, Arimaa, and offered the Arimaa Challenge Prize to promote breakthrough research in AI. Omar’s long-term vision is a world where everyone receives an unconditional basic income based on a stable cryptocurrency so that poverty and hunger are eliminated.
The Shardeum network will have a native coin called Shard with a ticker symbol of $SHM. The coin will be mined by validators, archive and standby nodes as reward for providing resources to the network. The coin will be used for paying gas fees associated with executing transfer transactions as well as smart contract execution on the Shardeum network.
Max supply: 508 million SHM Distribution
- 51% Community - reward to nodes; validators, archive and standby servers.
- 18% Sale - 3 month cliff then 2 years linear vesting.
- 15% Team - 3 month cliff then 2 years linear vesting.
- 11% Foundation - unlocked at Token Generation Event (TGE).
- 5% Ecosystem - unlocked at TGE.
As can be seen from the token allocation, the token allocation is quite acceptable for every stakeholder. The overall distribution mechanism for ecosystem participants is in place already but the specific reward schedule for node mining is not finalized yet. The project also looks to raise funds from venture firms, but those investors and the Shardeum team won't hold a large amount of tokens. And according to the official documents about tokenomics, SHM will be a deflationary coin. So far, it cannot predict exactly all metrics to initially evaluate the price and when the supply is completely diluted out the market.
By using the Shardus framework and adding EVM and smart contracting functionality at the application layer, Shardeum will be able to reach mainnet much faster than otherwise possible.
- Q2 2022 - Alphanet and Private Sale.
- Q3 2022 - Betanet
- Q4 2022 - Mainnet and Public Sale.
Thoughts On Shardeum
Shardeum pioneers a new approach with sharding. With dynamic state sharding on top of its EVM compatibility, Shardeum will be truly capable of handling multitude of DApps and layer 2 solutions across industries and consumers with optimum performance. It will indeed maneuver in unchartered territories in the blockchain space. Otherwise it will not be aiming for 1 million TPS with high decentralization and security. However, that only adds to the excitement to finally see another blockchain that is equal to or better than Ethereum and its smart contracts at some point in the near future.