Paper 2024/1315 Original: eprint.iacr.org/2024/1315 ↗

PulpFHE: Complex Instruction Set Extensions for FHE Processors

Omar Ahmed
Nektarios Georgios Tsoutsos
Abstract

The proliferation of attacks to cloud computing, coupled with the vast amounts of data outsourced to online services, continues to raise major concerns about the privacy for end users. Traditional cryptography can help secure data transmission and storage on cloud servers, but falls short when the already encrypted data needs to be processed by the cloud provider. An emerging solution to this challenge is fully homomorphic encryption (FHE), which enables computations directly on encrypted data, and recent works have focused on developing new processor designs tailored for native processing of FHE data. In this work, we introduce PulpFHE, an optimized instruction set extension tailored for the next generation of FHE processors. Our proposed FHE instructions offer native support for non-linear operations on encrypted data, and enable significantly faster homomorphic computations for a broad range of realistic applications.

Metadata
Available format(s)
PDF
Category
Applications
Publication info
Preprint.
History
2024-08-22: last updated
2024-08-22: published
Short URL
https://ia.cr/2024/1315
License
https://creativecommons.org/licenses/by/4.0/

BibTeX

@misc{cryptoeprint:2024/1315,
      author = {Omar Ahmed and Nektarios Georgios Tsoutsos},
      title = {PulpFHE: Complex Instruction Set Extensions for FHE Processors},
      howpublished = {Cryptology {ePrint} Archive, Paper 2024/1315},
      year = {2024},
      url = {https://eprint.iacr.org/2024/1315}
}
Note: In order to protect the privacy of readers, this archive does not use cookies or embedded third party content.