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
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
// SPDX-License-Identifier: AGPL-3.0-or-later
//! PRNG Xoshiro128** - High-quality pseudorandom f64 generator - Pure WGSL
//!
//! Deep Debt Principles:
//! - Self-knowledge: Operation knows its computation
//! - Zero hardcoding: Hardware-agnostic implementation
//! - Modern idiomatic Rust: Safe, zero unsafe code
//! - Complete implementation: Production-ready, no mocks
//! - Hardware-agnostic: Pure WGSL for universal compute
//!
//! f64 pipeline: seeds as `array<u32>` (1 per output, expanded to 4-stride),
//! output as `array<f64>` in [0, 1).
use crate::device::DeviceCapabilities;
use crate::error::Result;
use crate::tensor::Tensor;
/// Xoshiro128** PRNG for GPU random number generation.
pub struct PrngXoshiro {
seeds: Tensor,
offset: u32,
}
impl PrngXoshiro {
/// Creates a new PRNG with the given seeds tensor (u32) and offset.
#[must_use]
pub fn new(seeds: Tensor, offset: u32) -> Self {
Self { seeds, offset }
}
/// Xoshiro128** stateful PRNG (neuralSpring): per-thread state, `n_samples` per thread.
#[must_use]
pub fn wgsl_xoshiro128ss() -> &'static str {
static SHADER: std::sync::LazyLock<String> = std::sync::LazyLock::new(|| {
include_str!("../shaders/misc/xoshiro128ss_f64.wgsl").to_string()
});
std::sync::LazyLock::force(&SHADER).as_str()
}
/// WGSL kernel for Xoshiro PRNG (f32 variant).
pub const WGSL_PRNG_XOSHIRO_F32: &str = include_str!("../shaders/misc/prng_xoshiro.wgsl");
/// f64 version for universal math library portability.
#[must_use]
pub fn wgsl_shader_f64() -> &'static str {
include_str!("../shaders/misc/prng_xoshiro_f64.wgsl")
}
/// Executes the PRNG and returns random values in [0, 1).
///
/// # Errors
///
/// Returns [`Err`] if buffer allocation, GPU dispatch, or buffer
/// readback fails (e.g. device lost or out of memory).
pub fn execute(self) -> Result<Tensor> {
let device = self.seeds.device();
let seed_count: usize = self.seeds.shape().iter().product();
if seed_count == 0 {
return Ok(Tensor::new(vec![], vec![0], device.clone()));
}
// f64 shader expects 4 u32s per output (seed_base = idx * 4); expand 1→4 stride
let seeds_data = device.read_buffer_u32(self.seeds.buffer(), seed_count)?;
let expanded: Vec<u32> = (0..seed_count)
.flat_map(|i| [seeds_data[i], 0u32, 0u32, 0u32])
.collect();
let seeds_buffer = device
.device
.create_buffer_init(&wgpu::util::BufferInitDescriptor {
label: Some("PRNG Xoshiro seeds expanded"),
contents: bytemuck::cast_slice(&expanded),
usage: wgpu::BufferUsages::STORAGE,
});
let output_buffer = device.create_buffer_f64(seed_count)?;
#[repr(C)]
#[derive(Copy, Clone, bytemuck::Pod, bytemuck::Zeroable)]
struct Params {
size: u32,
offset: u32,
}
let params = Params {
size: seed_count as u32,
offset: self.offset,
};
let params_buffer = device
.device
.create_buffer_init(&wgpu::util::BufferInitDescriptor {
label: Some("PRNG Xoshiro Params"),
contents: bytemuck::cast_slice(&[params]),
usage: wgpu::BufferUsages::UNIFORM,
});
let bind_group_layout =
device
.device
.create_bind_group_layout(&wgpu::BindGroupLayoutDescriptor {
label: Some("PRNG Xoshiro f64 Bind Group Layout"),
entries: &[
wgpu::BindGroupLayoutEntry {
binding: 0,
visibility: wgpu::ShaderStages::COMPUTE,
ty: wgpu::BindingType::Buffer {
ty: wgpu::BufferBindingType::Storage { read_only: true },
has_dynamic_offset: false,
min_binding_size: None,
},
count: None,
},
wgpu::BindGroupLayoutEntry {
binding: 1,
visibility: wgpu::ShaderStages::COMPUTE,
ty: wgpu::BindingType::Buffer {
ty: wgpu::BufferBindingType::Storage { read_only: false },
has_dynamic_offset: false,
min_binding_size: None,
},
count: None,
},
wgpu::BindGroupLayoutEntry {
binding: 2,
visibility: wgpu::ShaderStages::COMPUTE,
ty: wgpu::BindingType::Buffer {
ty: wgpu::BufferBindingType::Uniform,
has_dynamic_offset: false,
min_binding_size: None,
},
count: None,
},
],
});
let bind_group = device.device.create_bind_group(&wgpu::BindGroupDescriptor {
label: Some("PRNG Xoshiro Bind Group"),
layout: &bind_group_layout,
entries: &[
wgpu::BindGroupEntry {
binding: 0,
resource: seeds_buffer.as_entire_binding(),
},
wgpu::BindGroupEntry {
binding: 1,
resource: output_buffer.as_entire_binding(),
},
wgpu::BindGroupEntry {
binding: 2,
resource: params_buffer.as_entire_binding(),
},
],
});
let shader = device.compile_shader_f64(Self::wgsl_shader_f64(), Some("PRNG Xoshiro f64"));
let pipeline_layout =
device
.device
.create_pipeline_layout(&wgpu::PipelineLayoutDescriptor {
label: Some("PRNG Xoshiro Pipeline Layout"),
bind_group_layouts: &[&bind_group_layout],
immediate_size: 0,
});
let pipeline = device
.device
.create_compute_pipeline(&wgpu::ComputePipelineDescriptor {
label: Some("PRNG Xoshiro Pipeline"),
layout: Some(&pipeline_layout),
module: &shader,
entry_point: Some("main"),
cache: None,
compilation_options: Default::default(),
});
let mut encoder = device.create_encoder_guarded(&wgpu::CommandEncoderDescriptor {
label: Some("PRNG Xoshiro Encoder"),
});
{
let mut pass = encoder.begin_compute_pass(&wgpu::ComputePassDescriptor {
label: Some("PRNG Xoshiro Pass"),
timestamp_writes: None,
});
pass.set_pipeline(&pipeline);
pass.set_bind_group(0, Some(&bind_group), &[]);
let caps = DeviceCapabilities::from_device(device);
let workgroups = caps.dispatch_1d(seed_count as u32);
pass.dispatch_workgroups(workgroups, 1, 1);
}
device.submit_commands(Some(encoder.finish()));
Ok(Tensor::from_buffer(
output_buffer,
self.seeds.shape().to_vec(),
device.clone(),
))
}
}
impl Tensor {
/// Generate random f64 values in [0, 1) using xoshiro128** PRNG.
/// Seeds tensor must contain u32 data (use `Tensor::from_data_pod` with u32).
///
/// # Errors
///
/// Returns [`Err`] if buffer allocation, GPU dispatch, or buffer
/// readback fails (e.g. device lost or out of memory).
pub fn prng_xoshiro(self, offset: u32) -> Result<Self> {
PrngXoshiro::new(self, offset).execute()
}
}
#[cfg(test)]
mod tests {
use super::*;
#[tokio::test]
async fn test_prng_xoshiro() {
let device = crate::device::test_pool::get_test_device().await;
// Seeds as u32 (one per output element)
let seeds: Vec<u32> = vec![1, 2, 3, 4, 5, 100, 200, 300];
let seeds_tensor = Tensor::from_data_pod(&seeds, vec![8], device).unwrap();
let output = seeds_tensor.prng_xoshiro(0).unwrap();
let result = output.to_f64_vec().unwrap();
assert_eq!(result.len(), 8);
assert!(
result
.iter()
.all(|&x| (0.0..1.0).contains(&x) && x.is_finite())
);
}
}