Public blockchains are the engine room of crypto. These digital ledgers record account balances, contract codes and other data using complex digital keys. Knowing that these records are public and cannot be deleted, modified, or copied engenders the trust that allows dispersed groups of collaborators to work or transact together on blockchains without the need for an intermediary. This trust is enhanced by duplicating and verifying information on multiple computers on a network. For this reason, many original blockchains cannot process more transactions than a single computer on the network can handle. This can lead to blockchains being overwhelmed with workload, causing delays and exorbitant costs for users, especially during periods of heavy crypto market activity. In September, Bitcoin was unable to handle more than around seven transactions per second and Ethereum, the second most popular cryptonet, was limited to around 15 per second – a lifespan compared to conventional electronic exchanges.
2. Why is it a trilemma?
Because extending a blockchain beyond a certain point inevitably compromises two of its fundamental characteristics: its decentralized structure, which gives it the transparency and trust of users to operate independently of third parties and governments, and its security ( protect data from hackers). In short, you can have “scalability”, decentralization, or security, but you can’t have all three.
3. Did anyone see this coming?
Yes. Computer scientist Hal Finney, who received the very first Bitcoin transaction from the token’s pseudonymous founder, Satoshi Nakamoto, pointed out early on that blockchains in their original design could not scale on their own. He proposed adding a simpler and more efficient secondary system on top of the main blockchain. “Bitcoin itself cannot scale so that every financial transaction in the world is broadcast to everyone and included on the blockchain,” Finney wrote in a forum in 2010. Ethereum co-founder Vitalik Buterin, coined the term “Blockchain trilemma” in 2017, stating the trade-offs needed to achieve “scalability”.
There have been several innovations to improve the performance of blockchains, but closer examination shows that they all weaken decentralization or security in the name of scalability. Here are some approaches:
• Larger Blocks: A blockchain is modified to bundle transactions into larger packets before they are validated and added to the network, improving its performance. This can be achieved by splitting a new blockchain from the original one in a process known as “forking”. Bitcoin Cash is one of the most important of these offshoots.
• New Layers: A protocol built on top of an existing blockchain that can handle transactions independently – something closer to what Finney was suggesting. Some examples of these so-called “Layer-2” protocols are Ethereum’s Polygon and Bitcoin’s Lightning Network.
• Sharding: Splitting chunks of data into smaller chunks to distribute the compute and storage workload across the network. Information in a partition can still be shared, which helps keep the network relatively decentralized and secure.
5. What is the impact of the trilemma?
This wasn’t a problem back when crypto was a niche technology used by a core group of enthusiasts. Now that traditional finance and other mainstream industries are turning to blockchains as a transparent and trusted environment for exchange and collaboration, these limitations are increasingly becoming a hindrance. Ethereum’s periodic congestion and high fees have caused it to lose market share in decentralized finance applications to rival blockchains such as Binance Smart Chain and Solana, which may be faster and cheaper because they can use fewer parties to order transactions. Between early 2021 and September 2022, Ethereum’s market share in DeFi, expressed in terms of total value locked, fell to 58% from 96%, according to data platform Defi Llama. Its backers hope to overcome these issues by changing the way the platform orders transactions.
More stories like this are available at bloomberg.com