Through a decades-long endeavor of building secure, robust and performant systems, we’ve developed a rich and deep understanding of centralized computing models. However, with recent advances in the area of blockchain, a new decentralized model gives rise to even stronger security guarantees. Cryptocurrencies and smart contracts are just two examples showcasing the promises of the blockchain model of computation. Concurrently, another fundamentally different approach to achieve stronger security is trusted execution environment (TEE), which has also seen a great advance recently with the debut of Intel SGX, a CPU-based implementation of TEE.
However, despite the nice features offered by TEE and blockchain, neither is ideal. The current blockchain systems suffer from serious practical limitations, e.g. poor performance, high energy consumption and lack of confidentiality. On the other hand, TEE is imperfect in its specification and implementation, and in isolation does not offer satisfactory availability guarantees. Motivated by these practical concerns, my research focuses on understanding the principles of a hybrid model that has the best of both worlds. In this talk, I will talk about Town Crier and Ekiden, two systems we built that demonstrate the benefits of synthesizing TEE and blockchains, and the pitfalls arising from harmonizing them.
Fan Zhang is a PhD candidate in the Department of Computer Science at Cornell University. Advised by Prof. Ari Juels, Fan’s research interest is in blockchains, trusted hardware, applied cryptography. Fan is particularly interested in understanding and building hybrid systems that enjoy the best of both blockchains and trusted hardware. Fan is a member of the Initiative for Cryptocurrencies & Contracts (IC3). Before joining Cornell, Fan earned his bachelor's degree in Electronic Engineering from Tsinghua University in 2014.