GPU acceleration

This guide explains how to update your existing program to leverage GPU acceleration, or to start a new program using GPU.

TFHE-rs now supports a GPU backend with CUDA implementation, enabling integer arithmetics operations on encrypted data.

Prerequisites

• Cuda version >= 10

• Compute Capability >= 3.0

• gcc >= 8.0 - check this page for more details about nvcc/gcc compatible versions

• cmake >= 3.24

• Rust version - check this page

Importing to your project

To use the TFHE-rs GPU backend in your project, add the following dependency in your Cargo.toml.

If you are using an x86 machine:

tfhe = { version = "0.6.4", features = [ "boolean", "shortint", "integer", "x86_64-unix", "gpu" ] }

If you are using an ARM machine:

tfhe = { version = "0.6.4", features = [ "boolean", "shortint", "integer", "aarch64-unix", "gpu" ] }

For optimal performance when using TFHE-rs, run your code in release mode with the --release flag.

Supported platforms

TFHE-rs GPU backend is supported on Linux (x86, aarch64).

OS	x86	aarch64
Linux	x86_64-unix	aarch64-unix*
macOS	Unsupported	Unsupported*
Windows	Unsupported	Unsupported

A first example

Configuring and creating keys.

Comparing to the CPU example, GPU set up differs in the key creation, as detailed here

Here is a full example (combining the client and server parts):

use tfhe::{ConfigBuilder, set_server_key, FheUint8, ClientKey, CompressedServerKey};
use tfhe::prelude::*;

fn main() {

    let config = ConfigBuilder::default().build();

    let client_key= ClientKey::generate(config);
    let compressed_server_key = CompressedServerKey::new(&client_key);

    let gpu_key = compressed_server_key.decompress_to_gpu();

    let clear_a = 27u8;
    let clear_b = 128u8;

    let a = FheUint8::encrypt(clear_a, &client_key);
    let b = FheUint8::encrypt(clear_b, &client_key);

    //Server-side

    set_server_key(gpu_key);
    let result = a + b;

    //Client-side
    let decrypted_result: u8 = result.decrypt(&client_key);

    let clear_result = clear_a + clear_b;

    assert_eq!(decrypted_result, clear_result);
}

Setting the keys

The configuration of the key is different from the CPU. More precisely, if both client and server keys are still generated by the client (which is assumed to run on a CPU), the server key has then to be decompressed by the server to be converted into the right format. To do so, the server should run this function: decompressed_to_gpu().

Once decompressed, the operations between CPU and GPU are identical.

Encryption

On the client-side, the method to encrypt the data is exactly the same than the CPU one, as shown in the following example:

    let clear_a = 27u8;
    let clear_b = 128u8;
    
    let a = FheUint8::encrypt(clear_a, &client_key);
    let b = FheUint8::encrypt(clear_b, &client_key);

Computation

The server first need to set up its keys with set_server_key(gpu_key).

Then, homomorphic computations are performed using the same approach as the CPU operations.

    //Server-side
    set_server_key(gpu_key);
    let result = a + b;

    //Client-side
    let decrypted_result: u8 = result.decrypt(&client_key);

    let clear_result = clear_a + clear_b;

    assert_eq!(decrypted_result, clear_result);

Decryption

Finally, the client decrypts the results using:

    let decrypted_result: u8 = result.decrypt(&client_key);

Improving performance.

TFHE-rs allows to leverage the high number of threads given by a GPU. To maximize the number of GPU threads, update your configuration accordingly:

let config = ConfigBuilder::with_custom_parameters(PARAM_GPU_MULTI_BIT_MESSAGE_2_CARRY_2_GROUP_3_KS_PBS, None).build();

Here's the complete example:

use tfhe::{ConfigBuilder, set_server_key, FheUint8, ClientKey, CompressedServerKey};
use tfhe::prelude::*;
use tfhe::shortint::parameters::PARAM_GPU_MULTI_BIT_MESSAGE_2_CARRY_2_GROUP_3_KS_PBS;

fn main() {

    let config = ConfigBuilder::with_custom_parameters(PARAM_GPU_MULTI_BIT_MESSAGE_2_CARRY_2_GROUP_3_KS_PBS, None).build();

    let client_key= ClientKey::generate(config);
    let compressed_server_key = CompressedServerKey::new(&client_key);

    let gpu_key = compressed_server_key.decompress_to_gpu();

    let clear_a = 27u8;
    let clear_b = 128u8;

    let a = FheUint8::encrypt(clear_a, &client_key);
    let b = FheUint8::encrypt(clear_b, &client_key);

    //Server-side

    set_server_key(gpu_key);
    let result = a + b;

    //Client-side
    let decrypted_result: u8 = result.decrypt(&client_key);

    let clear_result = clear_a + clear_b;

    assert_eq!(decrypted_result, clear_result);
}

List of available operations

The GPU backend includes the following operations:

name	symbol	Enc/Enc	Enc/ Int
Neg	-	✔️	N/A
Add	+	✔️	✔️
Sub	-	✔️	✔️
Mul	*	✔️	✔️
Div	/	✖️	✖️
Rem	%	✖️	✖️
Not	!	✔️	N/A
BitAnd	&	✔️	✔️
BitOr	|	✔️	✔️
BitXor	^	✔️	✔️
Shr	>>	✔️	✔️
Shl	<<	✔️	✔️
Rotate right	rotate_right	✔️	✔️
Rotate left	rotate_left	✔️	✔️
Min	min	✔️	✔️
Max	max	✔️	✔️
Greater than	gt	✔️	✔️
Greater or equal than	ge	✔️	✔️
Lower than	lt	✔️	✔️
Lower or equal than	le	✔️	✔️
Equal	eq	✔️	✔️
Cast (into dest type)	cast_into	✖️	N/A
Cast (from src type)	cast_from	✖️	N/A
Ternary operator	if_then_else	✔️	✖️

The equivalent signed operations are also available.

All operations follow the same syntax than the one described in here.

Benchmarks

All GPU benchmarks presented here were obtained on a single H100 GPU, and rely on the multithreaded PBS algorithm. The cryptographic parameters PARAM_GPU_MULTI_BIT_MESSAGE_2_CARRY_2_GROUP_3_KS_PBS were used.

The following table shows the performance when the inputs of the benchmarked operation are encrypted:

Operation \ Size	FheUint7	FheUint16	FheUint32	FheUint64	FheUint128	FheUint256
Negation (-)	46 ms	60 ms	75 ms	94 ms	150 ms	247 ms
Add / Sub (+,-)	46 ms	60 ms	75 ms	94 ms	150 ms	247 ms
Mul (x)	83 ms	121 ms	195 ms	456 ms	1.35 s	4.74 s
Equal / Not Equal (eq, ne)	25 ms	26 ms	38 ms	41 ms	52 ms	78 ms
Comparisons (ge, gt, le, lt)	46 ms	60 ms	74 ms	90 ms	109 ms	153 ms
Max / Min (max,min)	71 ms	86 ms	101 ms	124 ms	165 ms	236 ms
Bitwise operations (&, |, ^)	11 ms	12 ms	13 ms	15 ms	23 ms	32 ms
Left / Right Shifts (<<, >>)	71 ms	88 ms	109 ms	180 ms	279 ms	494 ms
Left / Right Rotations (left_rotate, right_rotate)	71 ms	88 ms	109 ms	180 ms	279 ms	494 ms

The following table shows the performance when the left input of the benchmarked operation is encrypted and the other is a clear scalar of the same size:

Operation \ Size	FheUint7	FheUint16	FheUint32	FheUint64	FheUint128	FheUint256
Add / Sub (+,-)	46 ms	60 ms	75 ms	94 ms	152 ms	251 ms
Mul (*)	67 ms	101 ms	149 ms	282 ms	727 ms	2.11 s
Equal / Not Equal (eq, ne)	26 ms	27 ms	27 ms	41 ms	45 ms	57 ms
Comparisons (ge, gt, le, lt)	29 ms	41 ms	54 ms	69 ms	87 ms	117 ms
Max / Min (max,min)	53 ms	65 ms	81 ms	102 ms	142 ms	200 ms
Bitwise operations (&, |, ^)	11 ms	13 ms	13 ms	15 ms	23 ms	32 ms
Left / Right Shifts (<<, >>)	11 ms	12 ms	13 ms	15 ms	23 ms	32 ms
Left / Right Rotations (left_rotate, right_rotate)	11 ms	12 ms	13 ms	15 ms	23 ms	32 ms