cubek-random 0.3.0-pre.2

CubeK: Random Number Generation
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95

use cubecl::{prelude::*, std::tensor::ViewMut};

use crate::{
    RandomFamily, lcg_step, taus_step_0, taus_step_1, taus_step_2, to_unit_interval_closed_open,
};

use super::{PrngArgs, PrngRuntime, random};

#[derive(CubeLaunch, CubeType)]
pub(crate) struct Uniform {
    lower_bound: f32,
    upper_bound: f32,
}

#[derive(Debug)]
struct UniformFamily;

impl RandomFamily for UniformFamily {
    type Runtime = Uniform;
}

#[cube]
impl PrngRuntime for Uniform {
    fn inner_loop<E: Numeric, N: Size>(
        args: Uniform,
        write_index_base: usize,
        n_invocations: u32,
        #[comptime] n_values_per_thread: usize,
        state_0: &mut u32,
        state_1: &mut u32,
        state_2: &mut u32,
        state_3: &mut u32,
        output: &mut ViewMut<'_, Vector<E, N>, usize>,
    ) {
        let lower_bound = args.lower_bound;
        let upper_bound = args.upper_bound;

        let scale = upper_bound - lower_bound;

        let mut output_vector = Vector::empty();

        let num_iterations = n_values_per_thread / N::value();
        #[unroll(num_iterations <= 8)]
        for vector_index in 0..num_iterations {
            // vectorization
            #[unroll]
            for i in 0..N::value() {
                *state_0 = taus_step_0(*state_0);
                *state_1 = taus_step_1(*state_1);
                *state_2 = taus_step_2(*state_2);
                *state_3 = lcg_step(*state_3);

                let int_random = *state_0 ^ *state_1 ^ *state_2 ^ *state_3;
                let f32_random = to_unit_interval_closed_open(int_random);

                let f32_uniform = f32_random * f32::cast_from(scale) + f32::cast_from(lower_bound);

                let uniform = E::cast_from(f32_uniform);

                output_vector.insert(i, uniform);
            }

            let write_index = vector_index * n_invocations as usize + write_index_base;

            output.write_checked(write_index, output_vector);
        }
    }
}

impl PrngArgs for Uniform {
    type Args = Self;

    fn args<R: Runtime>(self) -> UniformLaunch<R> {
        UniformLaunch::new(self.lower_bound, self.upper_bound)
    }
}

/// Pseudo-random generator with uniform distribution
pub fn random_uniform<R: Runtime>(
    client: &ComputeClient<R>,
    lower_bound: f32,
    upper_bound: f32,
    out: TensorBinding<R>,
    dtype: StorageType,
) -> Result<(), LaunchError> {
    random::<UniformFamily, R>(
        client,
        Uniform {
            lower_bound,
            upper_bound,
        },
        out,
        dtype,
    )
}