Comparative Review of AWS and Azure Confidential Computing Systems

Abstract

Confidential computing is an emerging field that focuses on ensuring the privacy and security of data during computation, addressing the increasing concerns with respect to data breaches and unauthorized access to sensitive information in cloud environments. The paper provides a comparative review of two prominent cloud platforms offering confidential computing solutions, namely, AWS Nitro Enclaves and Azure Confidential VMs with Intel SGX Enclaves. We analyse both platforms across multiple dimensions, such as their core components, attestation processes, communication mechanisms, and performance limitations. Additionally, we will evaluate practical factors such as regional availability, types of instances, and resource allocations. By analysing the unique features, advantages, and trade-offs of each platform, the paper aims to provide a clear understanding of why confidential computing is important in the context of increasing data privacy requirements and regulatory frameworks, such as GDPR and how these technologies enable secure and isolated execution of sensitive workloads on the cloud.