Skip to content

Systems catalogue

Papers behind the systems table (docs/systems.html): LLM frameworks, libraries, compilers, accelerators and alternatives. Files are named group-id_slug.

Folder: papers/references/catalogue/

0. CKKS LLM frameworks surveyed in the SoK

System Paper Year Library Hardware Latency Code
POLARIS SoK: Private LLM Inference using Approximate Homomorphic Encryption 2026 OpenFHE + FIDESlib A100 GPU 18 s (BERT-Tiny), 83 s (BERT-Mini), unamortized
THOR THOR: Secure transformer inference with homomorphic encryption 2025 Liberate.FHE A100 GPU 602 s / 128-token sequence code
CEP / EncryptedLLM (GPU-accelerated FHE GPT-2) EncryptedLLM: Privacy-Preserving Large Language Model Inference via GPU-Accelerated Fully Homomorphic Encryption 2024 Custom OpenFHE GPU backend A100 GPU ~100 s per generated token
CACHEMIR Cachemir: Fully Homomorphic Encrypted Inference of Generative Large Language Model with KV Cache 2026 Lattigo + Phantom A100 GPU 97 s time-per-output-token
CERIUM A scalable multi-GPU framework for encrypted large-model inference 2025 Custom multi-GPU CKKS compiler 1-8x A100/H100/B200 GPUs 134 s prefill (128 tokens, 8x B200); 66 s BERT-Base on A100
SplitPrompt Scaling up FHE-based Privacy-Preserving ML: Higher Throughput, Longer Inputs for LLama-3-8B (v1: Scaling up Privacy-Preserving ML: A CKKS Implementation of Llama-2-7B) 2026 HEaaN2 8x RTX 4090 GPUs 33-34 s per output token; 85 s summarization with 4096 tokens (128 encrypted)
ARION ARION: Attention-optimized transformer inference on encrypted data 2025 Lattigo 64-core CPU 9.3 s (Tiny) / 225 s (Base) amortized per sequence, batch 256
ELLMO ELLMo: Packing- and Depth-Aware Encrypted Transformer Inference 2026 OpenFHE + FIDESlib H200 GPU 8 s unamortized
RL (Rovida & Leporati) Transformer-based language models and homomorphic encryption: An intersection with bert-tiny 2024 OpenFHE Apple M1 Pro CPU ~672 s unamortized (32 tokens) code
TRICYCLE TRICYCLE: Private transformer inference with tricyclic encodings 2025 OpenFHE 128-core CPU 322 s unamortized (27 tokens)
EFLA (Encryption-Friendly LLM Architecture) Encryption-friendly LLM architecture 2025 HEaaN RTX 4090 GPU 27 s per layer code
PRITRAN Pritran: Privacy-preserving inference for transformer-based language models under fully homomorphic encryption 2026 OpenFHE 64-core CPU 398 s (Tiny) / ~2000 s (Mini) unamortized
EUSTON Euston: Efficient and User-Friendly Secure Transformer Inference with Non-Interactivity 2026 SEAL + Phantom RTX 6000 GPU 46 s (BERT-Base) / 136 s (GPT-2-1.5B) / 484 s (Llama-3-8B) amortized, batch 32
Powerformer Powerformer: Efficient and high-accuracy privacy-preserving language model with homomorphic encryption 2025 Liberate.FHE A100 GPU 345 s unamortized (128 tokens)
MOAI MOAI: Module-optimizing architecture for non-interactive secure transformer inference 2026 SEAL + Phantom H200 / A100 GPU 141 s BERT-Base amortized; 225 s Llama-3-8B decode (8 tokens)
NEXUS Secure transformer inference made non-interactive 2025 SEAL + Phantom 4x A100 GPUs 37 s BERT-Base amortized (batch 32); 52 s/token Llama-3-8B code
STIP STIP: Efficient and secure non-interactive transformer inference via compact packing 2026 SEAL + Phantom A100 GPU 64 s (BERT-Base) / 142 s (GPT-2-1.5B) / 407 s (Llama-3-8B) amortized
ENSI ENSI: Efficient non-interactive secure inference for large language models 2025 OpenFHE + Phantom A100 GPU 958 s TTFT (batch 32, 2048 tokens)
Polyformer Converting transformers to polynomial form for secure inference over homomorphic encryption 2024 HELayers A100 GPU 305 s (128 tokens next-word prediction)
SecretVoice 1.0 Secretvoice1.0: Non-interactive secure transformer inference 2026 Microsoft SEAL 1-core CPU 1488 s per encoder block (32 tokens)
RZW (encrypted LoRA split model) Practical secure inference algorithm for a fine-tuned large language model based on fully homomorphic encryption 2025 (see paper) (see paper) 1.61 s/token (reported)

1. Newer CKKS / FHE LLM frameworks (2026, not in the SoK)

System Paper Year Library Hardware Latency Code
Odin An Open-Source End-to-End FHE Implementation for Privacy-Preserving Llama 3 8B Inference 2026
THEMIS THEMIS: A Co-Designed System for Encrypted Transformer Inference 2026
ROSETTA ROSETTA: Efficient and Accurate Privacy-Preserving LLM Decoding via Hybrid CKKS/TFHE Evaluation 2026
Terrazzo (GL scheme on GPU) Terrazzo: Memory-Aware GPU Framework for Private Transformer Inference 2026
Zephyr Zephyr: GPU-Efficient Homomorphic Encryption for Privacy-Preserving Transformer Inference 2026
Error-propagation-aware scale design Error Propagation-Aware Scale Design for Efficient Homomorphic Encryption-Based LLM Inference 2026
ATLAS ATLAS: Automated Approximation of Transformers for Efficient Homomorphic Inference in One Hour 2026
HEAT HEAT: Faster Fully Homomorphic Inference via Approximations-Weights Co-Adaptation 2026
CGF-Softmax CGF-Softmax: A Cumulant-Based Softmax Reformulation for Efficient Inference under Homomorphic Encryption 2026
EncFormer EncFormer: Secure and Efficient Transformer Inference over Encrypted Data 2026

2. CKKS core and variants

System Paper Year Library Hardware Latency Code
CKKS (HEAAN) Homomorphic Encryption for Arithmetic of Approximate Numbers 2017 code
RNS-CKKS (full RNS variant) A Full RNS Variant of Approximate Homomorphic Encryption 2018
Conjugate-invariant CKKS Approximate Homomorphic Encryption over the Conjugate-invariant Ring 2018
Reduced-error CKKS (FLEXIBLEAUTO) Approximate Homomorphic Encryption with Reduced Approximation Error 2022 code
IND-CPA-D attack on CKKS On the Security of Homomorphic Encryption on Approximate Numbers 2021
IND-CPA-D secure CKKS (noise flooding / DP) Securing Approximate Homomorphic Encryption Using Differential Privacy 2022 code
Multi-key CKKS Efficient Multi-Key Homomorphic Encryption with Packed Ciphertexts with Application to Oblivious Neural Network Inference 2019 code
Threshold / multiparty CKKS (MHE) Multiparty Homomorphic Encryption from Ring-Learning-with-Errors 2021 code
Bit-CKKS / bootstrapping bits with CKKS Bootstrapping Bits with CKKS 2024
BLEACH (discrete CKKS) BLEACH: Cleaning Errors in Discrete Computations over CKKS 2024
General functional bootstrapping in CKKS General Functional Bootstrapping using CKKS 2025 code
Discrete CKKS for high-precision bootstrapping Leveraging Discrete CKKS to Bootstrap in High Precision 2025
Fourier-extension functional bootstrapping High-Precision Functional Bootstrapping for CKKS from Fourier Extension 2026

3. CKKS bootstrapping and operation advances

System Paper Year Library Hardware Latency Code
First CKKS bootstrapping Bootstrapping for Approximate Homomorphic Encryption 2018
Improved CKKS bootstrapping (Chen-Chillotti-Song) Improved Bootstrapping for Approximate Homomorphic Encryption 2019
Non-sparse-key bootstrapping (Bossuat et al.) Efficient Bootstrapping for Approximate Homomorphic Encryption with Non-Sparse Keys 2021 code
High-precision bootstrapping (minimax + inverse sine) High-Precision Bootstrapping of RNS-CKKS Homomorphic Encryption Using Optimal Minimax Polynomial Approximation and Inverse Sine Function 2021
META-BTS META-BTS: Bootstrapping Precision Beyond the Limit 2022
Hybrid key switching Better Bootstrapping for Approximate Homomorphic Encryption 2020
KLSS key switching Accelerating HE Operations from Key Decomposition Technique 2023
Fused plaintext-ciphertext matmul and bootstrapping Plaintext-Ciphertext Matrix Multiplication and FHE Bootstrapping: Fast and Fused 2024
Grafting (CKKS modulus engineering) Grafting: Decoupled Scale Factors and Modulus in RNS-CKKS 2024
SHIP bootstrapping SHIP: A Shallow and Highly Parallelizable CKKS Bootstrapping Algorithm 2025
PaCo bootstrapping PaCo: Bootstrapping for CKKS via Partial CoeffToSlot 2025
Large-interval evaluation Efficient Homomorphic Evaluation on Large Intervals 2022
Fast homomorphic Softmax Fast and Accurate Homomorphic Softmax Evaluation 2024

4. Alternative FHE schemes

System Paper Year Library Hardware Latency Code
Gentry-Lee matrix FHE (GL scheme) Fully Homomorphic Encryption for Matrix Arithmetic 2025
Gentry's FHE Fully Homomorphic Encryption Using Ideal Lattices 2009 code
Paillier (partially HE) Public-Key Cryptosystems Based on Composite Degree Residuosity Classes 1999 code
BGV (Leveled) Fully Homomorphic Encryption without Bootstrapping 2012 code
BFV Somewhat Practical Fully Homomorphic Encryption 2012 code
Brakerski scale-invariant FHE Fully Homomorphic Encryption without Modulus Switching from Classical GapSVP 2012
Revisited BGV/BFV for finite fields Revisiting Homomorphic Encryption Schemes for Finite Fields 2021 code
GSW Homomorphic Encryption from Learning with Errors: Conceptually-Simpler, Asymptotically-Faster, Attribute-Based 2013
FHEW FHEW: Bootstrapping Homomorphic Encryption in Less Than a Second 2015 code
TFHE (CGGI) TFHE: Fast Fully Homomorphic Encryption over the Torus 2016 code
TFHE programmable bootstrapping for DNNs Programmable Bootstrapping Enables Efficient Homomorphic Inference of Deep Neural Networks 2021 code
TFHE floating point TFHE Gets Real: An Efficient and Flexible Homomorphic Floating-Point Arithmetic 2025
FINAL (NTRU + LWE) FINAL: Faster FHE Instantiated with NTRU and LWE 2022 code
High-precision integer HE (CLPX) High-Precision Arithmetic in Homomorphic Encryption 2018
CHIMERA (scheme switching) CHIMERA: Combining Ring-LWE-based Fully Homomorphic Encryption Schemes 2020
PEGASUS (CKKS + FHEW switching) PEGASUS: Bridging Polynomial and Non-polynomial Evaluations in Homomorphic Encryption 2021 code
Efficient bootstrapping for GL matrix FHE Efficient Bootstrapping in Fully Homomorphic Encryption for Matrix Arithmetic 2026
Low-depth bootstrapping for matrix-native FHE Low-Depth Bootstrapping for Matrix-Native FHE 2026
Trace-Factored BigSwitch (GL keys) Trace-Factored BigSwitch for Matrix-Friendly FHE 2026

5. Interactive MPC and HE+MPC hybrids

System Paper Year Library Hardware Latency Code
Gazelle GAZELLE: A Low Latency Framework for Secure Neural Network Inference 2018
Delphi Delphi: A Cryptographic Inference Service for Neural Networks 2020 code
CrypTFlow2 CrypTFlow2: Practical 2-Party Secure Inference 2020 code
Cheetah Cheetah: Lean and Fast Secure Two-Party Deep Neural Network Inference 2022 code
THE-X THE-X: Privacy-Preserving Transformer Inference with Homomorphic Encryption 2022
Iron Iron: Private Inference on Transformers 2022
MPCFormer MPCFormer: Fast, Performant and Private Transformer Inference with MPC 2023 code
PUMA PUMA: Secure Inference of LLaMA-7B in Five Minutes 2023 code
CipherGPT CipherGPT: Secure Two-Party GPT Inference 2023
SIGMA SIGMA: Secure GPT Inference with Function Secret Sharing 2024 code
BOLT BOLT: Privacy-Preserving, Accurate and Efficient Inference for Transformers 2024 code
BumbleBee BumbleBee: Secure Two-party Inference Framework for Large Transformers 2025 code
Nimbus Nimbus: Secure and Efficient Two-Party Inference for Transformers 2024 code
SecFormer SecFormer: Fast and Accurate Privacy-Preserving Inference for Transformer Models via SMPC 2024 code
Ditto Ditto: Quantization-aware Secure Inference of Transformers upon MPC 2024 code
MPC-minimized secure LLM inference (Marill) MPC-Minimized Secure LLM Inference 2024
Centaur Centaur: Bridging the Impossible Trinity of Privacy, Efficiency, and Performance in Privacy-Preserving Transformer Inference 2025
CrypTen CrypTen: Secure Multi-Party Computation Meets Machine Learning 2021 code
SecretFlow-SPU SecretFlow-SPU: A Performant and User-Friendly Framework for Privacy-Preserving Machine Learning 2023 code
MP-SPDZ MP-SPDZ: A Versatile Framework for Multi-Party Computation 2020 code
SHAFT SHAFT: Secure, Handy, Accurate, and Fast Transformer Inference 2025
CRAfT CRAfT: Constant-Round Non-Polynomial Approximation and P2P Network for Secure Transformer Inference 2026

6. Trusted hardware, statistical and integrity-only alternatives

System Paper Year Library Hardware Latency Code
NVIDIA Confidential Computing (H100/Blackwell) NVIDIA Confidential Computing 2023
Confidential computing on H100: benchmark Confidential Computing on NVIDIA Hopper GPUs: A Performance Benchmark Study 2024
Apple Private Cloud Compute Private Cloud Compute: A new frontier for AI privacy in the cloud 2024
Intel TDX / AMD SEV-SNP Confidential VMs (Intel TDX, AMD SEV-SNP) 2021 code
Slalom Slalom: Fast, Verifiable and Private Execution of Neural Networks in Trusted Hardware 2019 code
Split-N-Denoise (SnD) Split-and-Denoise: Protect Large Language Model Inference with Local Differential Privacy 2024
InferDPT InferDPT: Privacy-Preserving Inference for Closed-box Large Language Model 2023
PrivateLoRA PrivateLoRA for Efficient Privacy Preserving LLM 2023
zkLLM zkLLM: Zero Knowledge Proofs for Large Language Models 2024 code
EZKL EZKL: zkSNARKs for ONNX models 2023 code

7. FHE libraries, GPU back-ends and compilers

System Paper Year Library Hardware Latency Code
OpenFHE OpenFHE: Open-Source Fully Homomorphic Encryption Library 2022 code
Microsoft SEAL Simple Encrypted Arithmetic Library - SEAL 2017 code
Lattigo Lattigo: A Multiparty Homomorphic Encryption Library in Go 2020 code
HEaaN / HEaaN2 HEaaN library 2018 code
HElib Algorithms in HElib 2014 code
IBM HELayers HeLayers: A Tile Tensors Framework for Large Neural Networks on Encrypted Data 2023 code
Phantom Phantom: A CUDA-Accelerated Word-Wise Homomorphic Encryption Library 2024 code
FIDESlib FIDESlib: A Fully-Fledged Open-Source FHE Library for Efficient CKKS on GPUs 2025 code
Liberate.FHE Liberate.FHE: A New FHE Library for Bridging the Gap between Theory and Practice 2023 code
Cheddar Cheddar: A Swift Fully Homomorphic Encryption Library Designed for GPU Architectures 2024
100x faster GPU bootstrapping Over 100x Faster Bootstrapping in Fully Homomorphic Encryption through Memory-centric Optimization with GPUs 2021
TFHE-rs TFHE-rs: Pure Rust implementation of TFHE 2022 code
Concrete ML Concrete ML: Privacy-Preserving ML with FHE 2022 code
Orion Orion: A Fully Homomorphic Encryption Framework for Deep Learning 2025 code
HEIR HEIR: Homomorphic Encryption Intermediate Representation 2023 code
DaCapo DaCapo: Automatic Bootstrapping Management for Efficient Fully Homomorphic Encryption 2024
Fhelipe A Tensor Compiler with Automatic Data Packing for Simple and Efficient Fully Homomorphic Encryption 2024 code
EVA / CHET EVA: An Encrypted Vector Arithmetic Language and Compiler for Efficient Homomorphic Computation 2020 code
TenSEAL TenSEAL: A Library for Encrypted Tensor Operations Using Homomorphic Encryption 2021 code
HEonGPU HEonGPU: a GPU-based Fully Homomorphic Encryption Library 1.0 2024 code
TensorFHE TensorFHE: Achieving Practical Computation on Encrypted Data Using GPGPU 2023

8. FHE hardware accelerators

System Paper Year Library Hardware Latency Code
F1 F1: A Fast and Programmable Accelerator for Fully Homomorphic Encryption 2021
CraterLake CraterLake: A Hardware Accelerator for Efficient Unbounded Computation on Encrypted Data 2022
BTS BTS: An Accelerator for Bootstrappable Fully Homomorphic Encryption 2022
ARK ARK: Fully Homomorphic Encryption Accelerator with Runtime Data Generation and Inter-Operation Key Reuse 2022
SHARP SHARP: A Short-Word Hierarchical Accelerator for Robust and Practical Fully Homomorphic Encryption 2023
Trebuchet TREBUCHET: Fully Homomorphic Encryption Accelerator for Deep Computation 2023
BASALISC BASALISC: Programmable Hardware Accelerator for BGV Fully Homomorphic Encryption 2023
Intel HERACLES HERACLES: 8192-way SIMD Programmable Scalable Fully-Homomorphic Encryption SoC 2026