TFHEpp is full Scracthed pure C++ Ver. of TFHE. TFHEpp is slightly(about 10%) faster than original TFHE implementation. In addition to that, THFEpp supports Circuit Bootstrapping and Private Boootstrapping many LUT. TFHEpp depends on AVX2 because we use SPQLIOS FMA. If you want run TFHEpp without AVX2, see spqlios++ branch. It include pure C++ implementation of SPQLIOS as header only library, but slow.
This code includes utf-8 identifiers like α. Therefore, Clang and GCC10 or later are primarily supported compilers. GCC9 is not supported.
The default parameter is 128 bit security. Please add -DUSE_80BIT_SECURITY=ON to use faster but less secure parameter.
Some environments which do not support AVX2 cannot use spqlios. Instead of spqlios, TFHEpp can use fftw3.
To use fftw3, install libfftw3-dev and add -DUSE_FFTW3=ON to the compile option.
Codes under thirdparties directory contain third-party libraries, Randen, Cereal, and SPQLIOS. See the corresponding directory to check the licenses.
TFHEpp uses this as a Cryptographically Secure Pseudo-Random Number Generator (CSPRNG). Original repository is here. I just removed some unnecessary codes, no modification.
cereal is a header-only C++11 serialization library. TFHEpp uses this to export ciphertexts and keys. Cereal is treated by git submodule.
SPQLIOS is the FFT library using AVX2 that is dedicated to the ring R[X]/(X^N+1) for N a power of 2. These codes come from experimental-tfhe. We just renamed instances to adapt with our codes.
This is the AVX512 version of SPQLIOS developed in MOSFHET. I confirmed that this is faster than SPQLIOS on Intel i5-11400.
FFTW is one of the most famous FFT libraries.
CAUTION: REDISTRIBUTING BINARY WHICH DEPENDS ON FFTW3 MEANS YOU AGREED WITH GPLv3 OR LATER.
Intel MKL is the library provided by Intel and including FFTW compatible interface for FFT.
We assume to install MKL by this procedure and already ran source /opt/intel/mkl/bin/mklvars.sh.
Add -DUSE_MKL to the compile option to use MKL
spqliox_aarch64 is the FFT library for aarch64 forked from SPQLIOS.
This is slightly faster than FFTW3(average 1ms).
This library requires xbyak_aarch64, and
to use this library, add -DUSE_SPQLIOX_AARCH64=on to the compile option.
| FFTW3 | spqliox_aarch64 |
|---|---|
| 15.801ms | 14.368ms |
Following Code measure how many time homomorphic NAND takes on your computer with TFHEpp.
git clone https://github.com/virtualsecureplatform/TFHEpp
cd TFHEpp
mkdir build
cd build
cmake .. -DENABLE_TEST=ON
make
ulimit -s unlimited
./test/nand
If you want to run semantically equivalent test on original TFHE, run below code.
git clone https://github.com/tfhe/tfhe.git --recursive
cd tfhe
mkdir build
cd build
cmake ../src -DENABLE_TESTS=on -DENABLE_NAYUKI_PORTABLE=off -DENABLE_NAYUKI_AVX=off -DENABLE_SPQLIOS_AVX=off -DENABLE_SPQLIOS_FMA=on -DCMAKE_BUILD_TYPE=optim
make
./test/test-gate-bootstrapping-spqlios-fma
If you have Docker on your system, this will do above on docker.
git clone https://github.com/virtualsecureplatform/TFHEpp
cd TFHEpp
docker build -t tfheppbench .
This is for TFHE-10ms. Because TFHE-10ms only supports 80-bit security parameter, this is not included in Dockerfile
git clone https://github.com/virtualsecureplatform/tfhe-10ms.git --recursive
cd tfhe-10ms
mkdir build
cd build
cmake ../src -DENABLE_TESTS=on -DENABLE_NAYUKI_PORTABLE=off -DENABLE_NAYUKI_AVX=off -DENABLE_SPQLIOS_AVX=off -DENABLE_SPQLIOS_FMA=on -DCMAKE_BUILD_TYPE=optim
make
./test/test-bootstrapping-fft-spqlios-fma
Here is the slides (in japanese). https://nindanaoto.github.io/
For the people who want to cite this library directly (may be in addition to VSP paper), I give a below example of bibtex citation.
@misc{TFHEpp,
author = {Kotaro Matsuoka},
title = {{TFHEpp: pure C++ implementation of TFHE cryptosystem}},
year = {2020},
howpublished = {\url{https://github.com/virtualsecureplatform/TFHEpp}}
}