Blockchain Scalability: Part One
- An Introduction to the Scalability Problem in Blockchain Technology -
Scalability in business, it's about expanding while remaining effective. In manufacturing this would involve improving efficiency of production, logistics, distribution etc. Increase supply to adjust to demand. This comes with greater costs and liability. What constantly emerges are trade-offs: independence vs reach, simplicity vs. capacity etc. Therefore, increasing scale is the challenge of balancing these factors. The goal is not only high production, but efficiency because otherwise it will be of no ultimate benefit. It can also be applied to technology and systems and their ability to expand reach while maintaining effective deliverance. Often the word has specific parameters in each industry. It is of crucial importance in blockchain because, while the technology may be revolutionary, contrary to popular belief among enthusiasts, it currently cannot actually handle the objective of mass adoption.
When evaluating efficiency, the key factor we seek to measure is throughput. Throughput is a measure of how much of something (product, data etc.) is able to pass through a system (organization, network etc.) in a set time. In business, this refers to how much can be produced and delivered to balance the rate at which stocks are depleted. For computers, it usually refers to data through a system and the rate the system can process it.
In financial systems, the important factors are the speed at which transactions can be processed and the time taken to confirm those transactions. Speed is measured by transactions per second (TPS). This metric is useful to highlight blockchain technology's shortcomings: Visa handles 1700TPS, while MasterCard claims the ability to handle up to 5000TPS. Bitcoin, on the other hand, only does 4-7TPS. Some others are Ethereum at 25TPS, Dash 35TPS, Litecoin ($LTC) 56TPS and Ripple ($XRP) at 1500TPS. Due to unique features of its protocol, Avalanche ($AVAX) theoretically has no upper limit but currently processes approximately, 4500TPS. Solana ($SOL) claims up to 50,000TPS. Most new generation POS layer-1s can deliver at least thousands of TPS. These numbers sound high, but to facilitate mass adoption they actually still need to be greatly improved. PayPal, which is a closed system that only settles between its own accounts, averages at 193TPS. Although they do far fewer than credit card companies, it is still way ahead of traditional POW blockchain networks.
The other factor is confirmation latency. In credit card networks and online payment services, this is less relevant. Being centralized systems, confirmation only takes a few seconds as the company assumes risk in guaranteeing ultimate settlement. In blockchain networks, however, transaction confirmation requires agreement throughout all the nodes. With few exceptions, blockchain networks take minutes or even an hour or more in certain circumstances, to finally confirm transactions. There are various approaches in development for managing these challenges going forward. A select few networks offer near instant confirmation, which is generally achieved by 'layer2-like' solutions being integrated into the consensus protocol.
For the purposes of blockchain functions not necessarily involving financial settlement, any interaction between participants can be seen as a transaction, that is to say an exchange of information of some sort. While important metrics, TPS and confirmation latency only measure transaction speed. It communicates something of the final goal, but it doesn't measure efficiency itself. Therefore, true scalability in blockchain technology is dependent on other underlying factors.
The challenge with scalability in blockchains is decentralization itself. All nodes have equal rights. Key to the problem is how efficiently all the data can be broadcast to all the nodes so that the whole network is in agreement (consensus) of the state of the network. Bitcoin in particular struggles with scalability precisely because of it's highly decentralized proof-of-work protocol. For transactions to be verified and written into blocks requires “work” by definition. To paraphrase in lay terms: If work is the fundamental measure of security, then efficiency (reduced work) will inherently undermine security – a “catch-22” situation that is a major challenge. Apart from the Taproot upgrade which batches transactions for greater efficiency, the most effective solution has been the Lightning Network, which is an opt-in layer-2 solution.
Proof-of-stake protocols were developed partly to tackle the scalability concern of pioneering blockchains and new consensus mechanisms developed for in. The subsequent developments have highlighted what is known as the “blockchain trilemma” of scalability vs. security vs. decentralization. This is the “trade-off” situation that emerges for blockchain networks. One of these factors will always be compromised in pursuit of improving the others. This is the fundamental challenge going forward for next gen layer-1s and some projects, such as Algorand ($ALGO), have developed primarily to explore and solve it effectively.
In an upcoming post, I will delve deeper into some of the specific challenges of blockchain network scalability and various solutions being developed to solve them.
Republished from original here.
(My posts are my opinion, based on easily accessible public information, and are not at all intended as investment advice)
