Skip to content
A Fully Homomorphic Encryption Brainfuck virtual machine
Branch: master
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Type Name Latest commit message Commit time
Failed to load latest commit information.
arcanevm Fixed NUFHE Context Dec 20, 2019
.gitignore Added number Dec 9, 2019
LICENSE Create LICENSE Dec 19, 2019 Merge pull request #10 from sobolevn/patch-2 Dec 21, 2019
requirements.txt Setting things up Dec 8, 2019


ArcaneVM is a virtual machine that can execute instructions encrypted with Fully Homomorphic Encryption. Basically, it's a Brainfuck interpreter that can execute encrypted brainfuck code.

ArcaneVM relies on the nufhe library for its FHE implementation.

Join the chat on Matrix


Fully Homomorphic Encryption (FHE) is a relatively new encryption technology that allows for computations on encrypted data. This allows us not only to encrypt the inputs but also encrypt the computations themselves. The output is also encrypted and when decrypted gives the plaintext result of the computation.

This is extremely powerful for people who not only have sensitive data, but also have computations that are sensitive.

Since the program essentially has zero context on what it is computing, it has to run through every possible branch of the program. On top of this, the underlying encryption schemes that make FHE possible are very computationally expensive.

What this means in the context of ArcaneVM and Brainfuck is that the more instructions a program has, and the larger the tape that it is working with, the slower the program is going to become. This is because at every cycle, the VM has to go through every possible instruction on every possible data cell.

As a result, ArcaneVM is very very very slow.

All integers in ArcaneVM are 8 bits.


Execute Brainfuck code

from arcanevm import run

run("++[>+++<-]>>", tape_size=5)


  • Python 3
  • PyOpenCL
  • nufhe


sudo apt install python3-pyopencl
pip3 install -r requirements.txt
You can’t perform that action at this time.