CraterLake: a hardware accelerator for efficient unbounded computation on encrypted data
Nikola Samardzic, Axel Feldmann, Aleksandar Krastev, Nathan Manohar, Nicholas Genise, Srinivas Devadas, Karim Eldefrawy, Chris Peikert, Daniel Sánchez
Abstract
Fully Homomorphic Encryption (FHE) enables offloading computation to untrusted servers with cryptographic privacy. Despite its attractive security, FHE is not yet widely adopted due to its prohibitive overheads, about 10,000× over unencrypted computation. Recent FHE accelerators have made strides to bridge this performance gap. Unfortunately, prior accelerators only work well for simple programs, but become inefficient for complex programs, which bring additional costs and challenges.
We present CraterLake, the first FHE accelerator that enables FHE programs of unbounded size (i.e., unbounded multiplicative depth). Such computations require very large ciphertexts (tens of MBs each) and different algorithms that prior work does not support well. To tackle this challenge, CraterLake introduces a new hardware architecture that efficiently scales to very large ciphertexts, novel functional units to accelerate key kernels, and new algorithms and compiler techniques to reduce data movement.
We evaluate CraterLake on deep FHE programs, including deep neural networks like ResNet and LSTMs, where prior work takes minutes to hours per inference on a CPU. CraterLake outperforms a CPU by gmean 4,600× and the best prior FHE accelerator by 11.2× under similar area and power budgets. These speeds enable realtime performance on unbounded FHE programs for the first time.
Ask Lune
Know this paper in minutes, not an afternoon.
Lune can read every page of it and the top-tier papers that cite it, and gives you its claims, its evidence and its limits, each one cited to the paper it comes from. No skimming, no AI slop.
45 top-tier papers have cited it since. Know what they found before your reviewers do.
Lune can draw on
the full paper · the top-tier papers that cite it · related work
Free to start. No credit card required.
Lineage
- GME: GPU-based Microarchitectural Extensions to Accelerate Homomorphic Encryption cites F1: A Fast and Programmable Accelerator for Fully Homomorphic Encryption
- GME: GPU-based Microarchitectural Extensions to Accelerate Homomorphic Encryption cites HEAX: An Architecture for Computing on Encrypted Data
- GME: GPU-based Microarchitectural Extensions to Accelerate Homomorphic Encryption cites ARK: Fully Homomorphic Encryption Accelerator with Runtime Data Generation and Inter-Operation Key Reuse
- GME: GPU-based Microarchitectural Extensions to Accelerate Homomorphic Encryption cites TensorFHE: Achieving Practical Computation on Encrypted Data Using GPGPU
- F1: A Fast and Programmable Accelerator for Fully Homomorphic Encryption cites Searching for MobileNetV3
- F1: A Fast and Programmable Accelerator for Fully Homomorphic Encryption cites HEAX: An Architecture for Computing on Encrypted Data
- F1: A Fast and Programmable Accelerator for Fully Homomorphic Encryption cites GAZELLE: A Low Latency Framework for Secure Neural Network Inference
- Poseidon: Practical Homomorphic Encryption Accelerator cites Fast Private Set Intersection from Homomorphic Encryption
- Poseidon: Practical Homomorphic Encryption Accelerator cites F1: A Fast and Programmable Accelerator for Fully Homomorphic Encryption
- Poseidon: Practical Homomorphic Encryption Accelerator cites HEAX: An Architecture for Computing on Encrypted Data
- Poseidon: Practical Homomorphic Encryption Accelerator cites ARK: Fully Homomorphic Encryption Accelerator with Runtime Data Generation and Inter-Operation Key Reuse
- ARK: Fully Homomorphic Encryption Accelerator with Runtime Data Generation and Inter-Operation Key Reuse cites F1: A Fast and Programmable Accelerator for Fully Homomorphic Encryption
- ARK: Fully Homomorphic Encryption Accelerator with Runtime Data Generation and Inter-Operation Key Reuse cites HEAX: An Architecture for Computing on Encrypted Data
- ARK: Fully Homomorphic Encryption Accelerator with Runtime Data Generation and Inter-Operation Key Reuse cites GAZELLE: A Low Latency Framework for Secure Neural Network Inference
- TensorFHE: Achieving Practical Computation on Encrypted Data Using GPGPU cites F1: A Fast and Programmable Accelerator for Fully Homomorphic Encryption
- TensorFHE: Achieving Practical Computation on Encrypted Data Using GPGPU cites HEAX: An Architecture for Computing on Encrypted Data
- TensorFHE: Achieving Practical Computation on Encrypted Data Using GPGPU cites ARK: Fully Homomorphic Encryption Accelerator with Runtime Data Generation and Inter-Operation Key Reuse
- BOLT: Privacy-Preserving, Accurate and Efficient Inference for Transformers cites F1: A Fast and Programmable Accelerator for Fully Homomorphic Encryption
- BOLT: Privacy-Preserving, Accurate and Efficient Inference for Transformers cites ARK: Fully Homomorphic Encryption Accelerator with Runtime Data Generation and Inter-Operation Key Reuse
- BOLT: Privacy-Preserving, Accurate and Efficient Inference for Transformers cites GAZELLE: A Low Latency Framework for Secure Neural Network Inference
Cited by top-tier papers45
- ARK: Fully Homomorphic Encryption Accelerator with Runtime Data Generation and Inter-Operation Key ReuseJongmin Kim, Gwangho Lee, Sangpyo Kim, Gina Sohn et al.MICRO 2022 · 160 citations
- BOLT: Privacy-Preserving, Accurate and Efficient Inference for TransformersQi Pang, Jinhao Zhu, Helen Möllering, Wenting Zheng et al.S&P 2024 · 149 citations
- Poseidon: Practical Homomorphic Encryption AcceleratorYinghao Yang, Huaizhi Zhang, Shengyu Fan, Hang Lu et al.HPCA 2023 · 109 citations
- TensorFHE: Achieving Practical Computation on Encrypted Data Using GPGPUShengyu Fan, Zhiwei Wang, Weizhi Xu, Rui Hou et al.HPCA 2023 · 90 citations
- GME: GPU-based Microarchitectural Extensions to Accelerate Homomorphic EncryptionKaustubh Shivdikar, Yuhui Bao, Rashmi Agrawal, Michael Tian Shen et al.MICRO 2023 · 46 citations
Builds on11
- Searching for MobileNetV3Andrew Howard, Ruoming Pang, Hartwig Adam, Quoc V. Le et al.ICCV 2019 · 9,163 citations
- GAZELLE: A Low Latency Framework for Secure Neural Network InferenceChiraag Juvekar, Vinod Vaikuntanathan, Anantha P. ChandrakasanUSENIX Security 2018 · 1,075 citations
- Fast Private Set Intersection from Homomorphic EncryptionHao Chen, Kim Laine, Peter RindalCCS 2017 · 446 citations
- F1: A Fast and Programmable Accelerator for Fully Homomorphic EncryptionNikola Samardzic, Axel Feldmann, Aleksandar Krastev, Srinivas Devadas et al.MICRO 2021 · 294 citations
- HEAX: An Architecture for Computing on Encrypted DataM. Sadegh Riazi, Kim Laine, Blake Pelton, Wei DaiASPLOS 2020 · 244 citations
Related papers
- A Tensor Compiler with Automatic Data Packing for Simple and Efficient Fully Homomorphic EncryptionAleksandar Krastev, Nikola Samardzic, Simon Langowski, Srinivas Devadas et al.PLDI 2024 · 23 citations
- Hydra: Scale-out FHE Accelerator Architecture for Secure Deep Learning on FPGAYinghao Yang, Xicheng Xu, Haibin Zhang, Jie Song et al.HPCA 2025 · 7 citations
- BitPacker: Enabling High Arithmetic Efficiency in Fully Homomorphic Encryption AcceleratorsNikola Samardzic, Daniel SánchezASPLOS 2024 · 19 citations
- Cheetah: Optimizing and Accelerating Homomorphic Encryption for Private InferenceBrandon Reagen, Wooseok Choi, Yeongil Ko, Vincent T. Lee et al.HPCA 2021 · 147 citations
- Orion: A Fully Homomorphic Encryption Framework for Deep LearningAustin Ebel, Karthik Garimella, Brandon ReagenASPLOS 2025 · 40 citations
