This is the implementation of the NTT in TensorFHE: Achieving Practical Computation on Encrypted Data Using GPGPU.
First pull this image docker pull suen0/tensorfhe:v0.1,
and then run docker run --rm --runtime=nvidia --gpus all -itd --name YOUR_CON_NAME IMAGE_ID bash
to build a container.
Run conda activate TensorFHE to enter the environment.
## There are six params for tensorNTT class.
## input is data for ntt.
## length is length for input.
## level is the level of the current ciphertext.
## isGPU indicates whether to enable GPU, the default is True.
## isTCU indicates whether to enable tensor cores on GPU, the default is True.
ntt = tensorNTT(input, length, batch, level, isGPU, isTCU)
## isStats indicates whether to count the throughput of the current operation,
## the default is False.
ntt.ntt(isStats)
Content type
Image
Digest
sha256:f666d8ee4…
Size
14.3 GB
Last updated
over 2 years ago
docker pull suen0/tensorfhe:v0.2