Performance Analysis of Scientific Computing Workloads on General Purpose TEEs

Akram, Ayaz; Giannakou, Anna; Akella, Venkatesh; Lowe-Power, Jason; Peisert, Sean

doi:10.1109/ipdps49936.2021.00115

Cited by 22 publications

(5 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Enabling secure processing in such a hostile environment requires careful SVM installation and a process to verify the installation (called attestation), which are described in this paper in detail. The performance penalty of the on-the-fly data encryption and decryption during memory writes and reads in the TEE is on the order of a few percent compared to processing without the encryption turned on [11]. This development provoked the second branch of research on protecting data in use and the rise of confidential computing-the use of TEE in protecting the data that are processed on untrusted hardware.…”

Section: Trusted Execution Environmentsmentioning

confidence: 99%

“…Later experiments confirmed these conclusions. Akram et al compared the SGX and SEV approaches on high-performance computing (HPC) benchmarks and concluded that SGX is not appropriate for HPC because it has a limited secure memory size and a complicated programming model, which leads to significant performance degradation compared to the unencrypted execution [11].…”

Section: Trusted Execution Environmentsmentioning

confidence: 99%

“…As indicated in the introduction, there are several previous studies that have analyzed the performance penalty of trusted execution environments in different settings (e.g., [11,17]). In addition to these studies, we have tested the performance of machine learning applications executed in the CoCos.ai environment.…”

Section: Performance Evaluationmentioning

confidence: 99%

See 2 more Smart Citations

Secure Multiparty Computation Using Secure Virtual Machines

Miladinović,

Milaković,

Vukasović

et al. 2024

Electronics

View full text Add to dashboard Cite

The development of new processor capabilities which enable hardware-based memory encryption, capable of isolating and encrypting application code and data in memory, have led to the rise of confidential computing techniques that protect data when processed on untrusted computing resources (e.g., cloud). Before confidential computing technologies, applications that needed data-in-use protection, like outsourced or secure multiparty computation, used purely cryptographic techniques, which had a large negative impact on the processing performance. Processing data in trusted enclaves protected by confidential computing technologies promises to protect data-in-use while possessing a negligible performance penalty. In this paper, we have analyzed the state-of-the-art in the field of confidential computing and present a Confidential Computing System for Artificial Intelligence (CoCoS.ai), a system for secure multiparty computation, which uses virtual machine-based trusted execution environments (in this case, AMD Secure Encrypted Virtualization (SEV)). The security of the proposed solution, as well as its performance, have been formally analyzed and measured. The paper reveals many gaps not reported previously that still exist in the current confidential computing solutions for the secure multiparty computation use case, especially in the processes of creating new secure virtual machines and their attestation, which are tailored for single-user use cases.

show abstract

Section: Trusted Execution Environmentsmentioning

confidence: 99%

Section: Trusted Execution Environmentsmentioning

confidence: 99%

See 1 more Smart Citation

Secure Multiparty Computation Using Secure Virtual Machines

Miladinović,

Milaković,

Vukasović

et al. 2024

Electronics

View full text Add to dashboard Cite

show abstract

“…If they are all correct, the algorithm continues to be executed, and the intersection can be restored according to the results. (4) Step (5) uses the additive homomorphism of the Lifted-ElGamal threshold cryptosystem, i.e.,…”

Section: Correctness Analysismentioning

confidence: 99%

“…Cramer [3] also predicted that MPC would become a new and powerful tool for privacy protection. Cryptologists have studied the MPC algorithms in various fields including secure scientific computing solutions [1,[4][5][6], secure data mining [7,8], secure statistical analysis [9][10][11], secure computational geometry [12][13][14][15][16], and secure computing set [17][18][19][20]. These studies promoted the development of MPC.…”

Section: Introductionmentioning

confidence: 99%

Secure Multi-Party Computation of Graphs’ Intersection and Union under the Malicious Model

Liu

Luo³

et al. 2023

Electronics

View full text Add to dashboard Cite

In recent years, with the development of information security, secure multi-party computation has gradually become a research hotspot in the field of privacy protection. The intersection and union computation of graphs is an important branch of secure computing geometry. At present, the intersection and union of graphs are almost designed under the semi-honest model, and few solutions are proposed under the malicious model. However, the solution under the malicious model is more secure and has important theoretical and practical significance. In this paper, the possible malicious behaviors of computing the intersection and union of graphs are analyzed. Using the Lifted-ElGamal threshold cryptosystem and zero-knowledge proof method, the secure multi-party computation algorithm of graphs’ intersection and union under the malicious model is designed. The real/ideal model paradigm is used to prove the security of the algorithm, the efficiency of the algorithm is analyzed in detail, and the feasibility is verified through experiment.

show abstract