Accelerating NTRU based homomorphic encryption using GPUs

By Wei Dai, Yarkin Doröz, Berk Sunar. 2014

71 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Information Systems

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

1. Wei Dai
2. Kim Laine
3. M. Sadegh Riazi
4. Blake Pelton
5. Berk Sunar
6. Yarkin Doröz
7. Erkay Savas
8. Erdinç Öztürk

In Information Systems

1. Wei Dai
2. Kim Laine
3. M. Sadegh Riazi
4. Blake Pelton
5. Yarkin Doröz
6. Berk Sunar
7. Erdinç Öztürk
8. Erkay Savas

Citations in the corpus

• HEAX: An Architecture for Computing on Encrypted Data (2019, background): “Note that parameter sets corresponding to 2 11 (or lower) are almost never used in practice due to the multiplication depth of 1 (or zero).”
• A Survey on Homomorphic Encryption Schemes (2017, background): “While one of them focuses on a customized and a generic bit-sliced implementation of NTRU-like FHE schemes [Doröz et al. 2014] and the other suggests the use of GPU [Dai et al. 2014].”
• Accelerating Homomorphic Evaluation on Reconfigurable Hardware (2015, background): “These may be reasons that previous work mainly focuses on using GPUs [19,40,41] and ASICs [21,44], and that FPGA implementations either work only with small parameters and on-chip memory [11] or explicitly do not take into account the complexity of transferring data between an FPGA and a host [9,35]”
• Efficient Deep Learning on Multi-Source Private Data (2018, background): “The benefits of such cryptographic methods are that the root of trust is in the algorithms, themselves, and that they may be used directly with hardware accelerators [12, 7].”
• Accelerating LTV Based Homomorphic Encryption in Reconfigurable Hardware (2015, methodology): “[14] reports GPU acceleration for NTRU based FHE evaluating Prince and AES block ciphers, with 2.”
• cuHE: A Homomorphic Encryption Accelerator Library (2015, methodology): “Prince at 51 msec (1 GPU), using the cuHE library which is 25 times faster than the previously reported fastest implementation [9].”
• Accelerating Somewhat Homomorphic Evaluation using FPGAs (2015, methodology): “Mul Speedup Relin Speedup AES Speedup (msec) (sec) (sec) CPU [16] 970 1× 103 1× 55 1× GPU [14] 340 2.”
• A Custom Accelerator for Homomorphic Encryption Applications (2017, background): “Dor€oz et al. implemented Prince encryption algorithm which is more suitable for homomorphic encryption with low depth circuit and less computation complexity in [12] with a runtime of 57 minutes.”
• HEAX: High-Performance Architecture for Computation on Homomorphically Encrypted Data in the Cloud (2019, background): “Several works [58, 24, 27, 59, 20, 21] focus on improving the performance of YASHE [9] and LTV [53] schemes or their variants.”