As machine learning (ML) technologies and applications are rapidly changing many domains of computing, security issues associated with ML are also emerging. In the domain of systems security, many endeavors have been made to ensure ML model and data confidentiality. ML computations are often inevitably performed in untrusted environments and entail complex multi-party security requirements. Hence, researchers have leveraged the Trusted Execution Environments (TEEs) to build confidential ML computation systems. This paper conducts a systematic and comprehensive survey by classifying attack vectors and mitigation in TEE-protected confidential ML computation in the untrusted environment, analyzes the multi-party ML security requirements, and discusses related engineering challenges.

Demand for data-intensive workloads and confidential computing are the prominent research directions shaping the future of cloud computing. Computer architectures are evolving to accommodate the computing of large data. Meanwhile, a plethora of works has explored protecting the confidentiality of the in-cloud computation in the context of hardware-based secure enclaves. However, the approach has faced challenges in achieving efficient large data computation. In this paper, we present a novel design, called SE-PIM, that retrofits Processing-In-Memory (PIM) as a data-intensive confidential computing accelerator. PIM-accelerated computation renders large data computation highly efficient by minimizing data movement. Based on our observation that moving computation closer to memory can achieve efficiency of computation and confidentiality of the processed information simultaneously, we study the advantages of confidential computing inside memory. We construct our findings into a software-hardware co-design called SE-PIM. Our design illustrates the advantages of PIM-based confidential computing acceleration. We study the challenges in adapting PIM in confidential computing and propose a set of imperative changes, as well as a programming model that can utilize them. Our evaluation shows SE-PIM can provide a side-channel resistant secure computation offloading and run data-intensive applications with negligible performance overhead compared to the baseline PIM model.

Demand for data-intensive workloads and confidential computing are the prominent research directions shaping the future of cloud computing. Computer architectures are evolving to accommodate the computing of large data better. Protecting the computation of sensitive data is also an imperative yet challenging objective; processor-supported secure enclaves serve as the key element in confidential computing in the cloud. However, side-channel attacks are threatening their security boundaries. The current processor architectures consume a considerable portion of its cycles in moving data. Near data computation is a promising approach that minimizes redundant data movement by placing computation inside storage. In this paper, we present a novel design for Processing-In-Memory (PIM) as a dataintensive workload accelerator for confidential computing. Based on our observation that moving computation closer to memory can achieve efficiency of computation and confidentiality of the processed information simultaneously, we study the advantages of confidential computing inside memory. We then explain our security model and programming model developed for PIMbased computation offloading. We construct our findings into a software-hardware co-design, which we call PIM-Enclave. Our design illustrates the advantages of PIM-based confidential computing acceleration. Our evaluation shows PIM-Enclave can provide a side-channel resistant secure computation offloading and run data-intensive applications with negligible performance overhead compared to baseline PIM model.