runmat-accelerate 0.4.5

Pluggable GPU acceleration layer for RunMat (CUDA, ROCm, Metal, Vulkan/Spir-V)
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
96
97
98
99
100
101
102
103

#[derive(Clone, Copy, Debug, PartialEq, Eq)]
pub enum NumericPrecision {
    F32,
    F64,
}

impl NumericPrecision {
    pub fn as_str(&self) -> &'static str {
        match self {
            NumericPrecision::F32 => "f32",
            NumericPrecision::F64 => "f64",
        }
    }
}

#[derive(Clone, Copy)]
pub enum BinaryOpCode {
    Add = 0,
    Sub = 1,
    Mul = 2,
    Div = 3,
    Hypot = 4,
    Atan2 = 5,
    Pow = 6,
    Max = 7,
    Min = 8,
}

#[derive(Clone, Copy)]
pub enum UnaryOpCode {
    Sin = 0,
    Cos = 1,
    Abs = 2,
    Exp = 3,
    Log = 4,
    Sqrt = 5,
    Sign = 6,
    Real = 7,
    Imag = 8,
    Conj = 9,
    Angle = 10,
    Expm1 = 11,
    Log1p = 12,
    Log10 = 13,
    Log2 = 14,
    Pow2 = 15,
    Floor = 16,
    Ceil = 17,
    Fix = 18,
    Tan = 19,
    Asin = 20,
    Acos = 21,
    Atan = 22,
    Sinh = 23,
    Cosh = 24,
    Tanh = 25,
    Asinh = 26,
    Acosh = 27,
    Atanh = 28,
    Gamma = 29,
    Factorial = 30,
    Single = 31,
    NextPow2 = 32,
}

#[derive(Clone, Copy)]
pub enum ScalarOpCode {
    Add = 0,
    Sub = 1,
    Mul = 2,
    Div = 3,
    RSub = 4,
    RDiv = 5,
    Max = 6,
    Min = 7,
}

#[derive(Clone, Copy)]
pub enum GlobalReduceOp {
    Sum = 0,
    Prod = 1,
    Min = 2,
    Max = 3,
    CountNonZero = 4,
}

#[derive(Clone, Copy)]
pub enum DimReduceOp {
    Sum = 0,
    Mean = 1,
    Prod = 2,
    AnyInclude = 3,
    AnyOmit = 4,
    AllInclude = 5,
    AllOmit = 6,
    CountNonZero = 7,
}

#[derive(Clone, Copy)]
pub enum DimReduceExtrema {
    Min = 0,
    Max = 1,
}