Can SegWit Save Proof of Work?

If you've been following news surrounding digital currencies and blockchain, you’ve probably seen the criticisms of the inefficiencies in Bitcoin's underlying blockchain architecture. Bitcoin relies on a computationally intensive consensus algorithm, proof of work (POW), to digitally establish the security and trust provided by central institutions in traditional networks. The system is designed to be transparent, so anyone in the network can independently verify the validity of a transaction and ensure their counterparty isn't trying to generate currency out of thin air or spend the same coin twice.

So far, POW has done its job. As of early 2018, more than 310 million transactions have been made on the Bitcoin blockchain alone without issue. But using math rather than institutions to back a currency requires time, money, and a lot of energy. These shortcomings call into question whether POW has a future in use cases—including many in the electricity industry—that require a faster, more efficient blockchain architecture.

In February 2018, the Bitcoin community announced long-awaited changes to the digital currency's core protocols that are aimed squarely at these inefficiencies. But segregated witness (SegWit) isn’t likely to save POW. Industries outside of the cryptocurrency world will need a new design altogether.

When Proof of Work Doesn't Work                                                               

POW is essentially a mathematical race that requires a lot of time and energy to complete; this limits the speed at which Bitcoin transactions can be confirmed to between 3 and 7 transactions per second (TPS). Ethereum, which also uses POW, manages about 15 TPS. This makes both networks highly sensitive to network congestion. Throughput capacity acts like a bottleneck in the system, and can cause a pileup if the number of transactions exceeds the rate at which they can be confirmed.

When network congestion occurs, users have to choose between offering a higher percentage of their transaction as a fee to validators, who are incentivized to confirm high value transactions first, or waiting even longer to be sure their transaction is finalized. The figures below display that Bitcoin transaction fees jumped to $55 per transaction in December 2017, while average wait times reached almost 20 hours.

Bitcoin Average Transaction Fee (Last 6 Months)

(Source: Bitinfocharts.com)

Bitcoin Average Transaction Confirmation Time (Last 6 Months)

(Source: Blockchain.info)

SegWit to the Rescue?

SegWit has complex technical details. The new protocol changes how information is structured in the blockchain database and allows more transactions to fit into a block, reducing the likelihood of congestion in the network. While congestion is still possible and fees could still spike, SegWit should help buffer against the magnitude of swings (such as those in 2017) and create more stability in the system.

So far, SegWit seems to be helping to keep fees low. However, while it makes the system more stable on average, it doesn't prevent congestion from happening—it only raises the threshold where it becomes a problem. And SegWit doesn't address other concerns with POW, like its massive energy consumption.

POW blockchains can support other applications, as Ethereum's range of decentralized applications shows. However, any business models that require rapid transaction confirmation or rely on low value transactions would be wise to consider a permissioned blockchain architecture or similar design that prioritizes throughput capacity over maximizing decentralization.

Many of the most transformative applications for blockchain rely on low volume transactions, and they will require an architecture that keeps such exchanges reliable and profitable. SegWit is a good stopgap, but POW-based blockchain architectures will have to do better if they hope to have value outside of the world of digital currencies.