Blockchains and Their Limits

Blockchain-based cryptocurrencies, led by Bitcoin, have emerged as potential new infrastructure for pseudonymous online payments, cheap remittance, trustless digital asset exchange, and smart contracts.

Yet, despite the excitement and euphoria that surrounds Bitcoin-derived blockchain protocols, the underlying technology suffers from fundamental limitations. This talk will tackle two of these limits, stemming from the consensus layer.

First, we will examine how malicious participants can game the Bitcoin protocol to earn more rewards than their fair share. We will see that Satoshi's protocol is not incentive-compatible: for the protocol to work as intended, the fraction of misbehaving nodes must not exceed 1/3rd, a result analogous to Lamport's 3f+1 result for strict consensus.

Second, we will focus on the inherent performance limitations of blockchain protocols. These limits have kicked off much acrimonious debate within the Bitcoin community, centered around how to adjust existing parameters to improve throughput. We will discuss how a re-thinking of the consensus layer can retain Bitcoin's open architecture, and simultaneously improve throughput and reduce latency by a few orders of magnitude, to the maximum achievable on the underlying network.

We will end with a brief discussion of the challenges facing blockchains.

Emin Gün Sirer is an associate professor in the Computer Science Department at Cornell University, and a co-Director of the Initiative for Cryptocurrencies and Smart Contracts. He has been working on peer-to-peer digital currencies based on proofs of work going back to 2002. With Ittay Eyal, he was the first to disprove some common folk theorems about the security of Bitcoin. He has, since then, played a key role in developing new techniques for improving blockchains.