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
// SPDX-License-Identifier: Apache-2.0
// Copyright (c) 2026 Fábio Henrique de Lima Silva (fhl.bsb@gmail.com) All rights reserved.
//! Scalar Reference Implementations for DSP math kernels.
//!
//! # Purpose
//!
//! The role of this module is to be the **scalar reference implementation**: simple,
//! correct, unoptimized algorithms that serve as an oracle in parity tests
//! for AVX2 and AVX-512 kernels. Every vector kernel must produce the same
//! numerical result (within floating-point tolerance) as its scalar counterpart here.
//!
//! # Source of Truth
//!
//! The definitive mathematical specification of each operation is NeuralAmpModelerCore (C++).
//! The implementations here are faithful scalar translations of that reference code,
//! used exclusively for internal validation.
// Re-exports of Wavenet fallbacks (Test Only).
pub use crate;
/// Convolution fallback implementations.
/// Dot product fallback implementations.
/// GEMM and GEMV fallback implementations.
/// LSTM gate fallback implementations.
/// Miscellaneous utility and helper fallback implementations.
pub use *;
pub use *;
pub use *;
pub use *;
pub use *;