A vision for autonomous blockchains backed by secure hardware

Abstract

Blockchains have emerged as a potential mechanism to enable immutable and consistent sharing of data across organizational boundaries. While much of the discussion on blockchains to date has been structured around public versus permissioned blockchains, both of these architectures have significant drawbacks. Public blockchains are energy inefficient, hard to scale and suffer from limited throughput and high latencies, while permissioned blockchains depend on specially designated nodes, potentially leak meta-information, and also suffer from scale and performance bottlenecks. This raises the question if blockchains, in their current form, are the only class of datastores that can provide such strong integrity guarantees. We introduce autonomous blockchains, an architecture based on free-standing, immutable, eidetic databases that implement independent timelines, linked together through interactions. Autonomous blockchains can be realized using trusted execution environments in combination with audit mechanisms. This architecture does not only provide blockchain-like integrity and auditability guarantees but also supports storing and querying private data. Further, multiple autonomous blockchains can be linked together through federated transactions to exchange data and order mutual operations. These transactions are amenable to audits and yield tamper-proof witnesses. Evaluation shows that this design can achieve high throughput while providing stronger integrity guarantees than conventional datastores.