Beyond Theory: Investigating the Practical Feasibility of Confidential Computing

Confidential Computing (CC) is a fairly new concept encompassing technologies designed to safeguard data during processing, playing a crucial role in data protection throughout its lifecycle. In this paper, we address the problem of ensuring data security from a technical perspective, delving into the intricate workings of technologies enabling CC. While the initial sections focus on establishing fundamental concepts related to CC, including Trusted Execution Environments (TEEs), later parts of the paper present practical prototypes demonstrating the setup of a CC environment on Cloud Computing Platforms (CCPs), such as Microsoft Azure. We specifically showcase two leading technologies in this domain: Intel Software Guard Extensions (SGX) and AMD Secure Encrypted Virtualization (SEV), both harnessing dedicated Central Processing Units (CPUs) from their respective manufacturers. This paper extensively covers the diverse approaches these technologies employ to secure data in use. Additionally, we offer a brief comparative analysis of related technologies as alternatives. Our research not only addresses the problem statement and motivation but also provides insights into significant results achieved in the realm of CC.