1use crate::{CubeRuntime, FloatElement, IntElement, element::BoolElement, tensor::CubeTensor};
2use burn_tensor::{
3 TensorData,
4 backend::{Backend, DeviceOps},
5};
6use cubecl::server::ComputeServer;
7use std::marker::PhantomData;
8
9#[cfg(not(feature = "fusion"))]
10use burn_ir::{BackendIr, TensorHandle};
11#[cfg(not(feature = "fusion"))]
12use burn_tensor::ops::{BoolTensor, FloatTensor, IntTensor, QuantizedTensor};
13
14#[derive(new)]
16pub struct CubeBackend<R: CubeRuntime, F: FloatElement, I: IntElement, BT: BoolElement> {
17 _runtime: PhantomData<R>,
18 _float_elem: PhantomData<F>,
19 _int_elem: PhantomData<I>,
20 _bool_elem: PhantomData<BT>,
21}
22
23impl<R, F, I, BT> Backend for CubeBackend<R, F, I, BT>
24where
25 R: CubeRuntime,
26 R::Server: ComputeServer,
27 R::Device: burn_tensor::backend::DeviceOps,
28 F: FloatElement,
29 I: IntElement,
30 BT: BoolElement,
31{
32 type Device = R::Device;
33
34 type FloatElem = F;
35 type IntElem = I;
36 type BoolElem = BT;
37
38 type FloatTensorPrimitive = CubeTensor<R>;
39 type IntTensorPrimitive = CubeTensor<R>;
40 type BoolTensorPrimitive = CubeTensor<R>;
41 type QuantizedTensorPrimitive = CubeTensor<R>;
42
43 fn name(device: &Self::Device) -> String {
44 let client = R::client(device);
45 format!("cubecl<{}>", R::name(&client))
46 }
47
48 fn seed(_device: &Self::Device, seed: u64) {
49 cubecl::random::seed(seed);
50 }
51
52 fn ad_enabled() -> bool {
53 false
54 }
55
56 fn sync(device: &Self::Device) {
57 let client = R::client(device);
58 futures_lite::future::block_on(client.sync());
59 }
60
61 fn memory_persistent_allocations<Output, Input, Func: Fn(Input) -> Output>(
62 device: &Self::Device,
63 input: Input,
64 func: Func,
65 ) -> Output {
66 let client = R::client(device);
67 client.memory_persistent_allocation(input, func)
68 }
69
70 fn memory_cleanup(device: &Self::Device) {
71 let client = R::client(device);
72 client.memory_cleanup();
73 }
74
75 fn staging<'a, Iter>(data: Iter, device: &Self::Device)
76 where
77 Iter: Iterator<Item = &'a mut TensorData>,
78 {
79 let client = R::client(device);
80 client.staging(data.map(|td| &mut td.bytes), false);
81 }
82}
83
84impl<R: CubeRuntime, F: FloatElement, I: IntElement, BT: BoolElement> core::fmt::Debug
85 for CubeBackend<R, F, I, BT>
86{
87 fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
88 f.write_str("CubeCLBackend")
89 }
90}
91
92impl<R: CubeRuntime, F: FloatElement, I: IntElement, BT: BoolElement> Clone
93 for CubeBackend<R, F, I, BT>
94{
95 fn clone(&self) -> Self {
96 Self::new()
97 }
98}
99
100impl<R: CubeRuntime, F: FloatElement, I: IntElement, BT: BoolElement> Default
101 for CubeBackend<R, F, I, BT>
102{
103 fn default() -> Self {
104 Self::new()
105 }
106}
107
108impl<R: cubecl::Runtime> CubeRuntime for R
109where
110 R::Device: DeviceOps,
111{
112 type CubeDevice = R::Device;
113 type CubeServer = R::Server;
114}
115
116#[cfg(not(feature = "fusion"))]
117impl<R: CubeRuntime, F: FloatElement, I: IntElement, BT: BoolElement> BackendIr
118 for CubeBackend<R, F, I, BT>
119{
120 type Handle = CubeTensor<R>;
121
122 fn float_tensor(handle: TensorHandle<Self::Handle>) -> FloatTensor<Self> {
123 handle.handle
124 }
125
126 fn int_tensor(handle: TensorHandle<Self::Handle>) -> IntTensor<Self> {
127 handle.handle
128 }
129
130 fn bool_tensor(handle: TensorHandle<Self::Handle>) -> BoolTensor<Self> {
131 handle.handle
132 }
133
134 fn quantized_tensor(handle: TensorHandle<Self::Handle>) -> QuantizedTensor<Self> {
135 handle.handle
136 }
137
138 fn float_tensor_handle(tensor: FloatTensor<Self>) -> Self::Handle {
139 tensor
140 }
141
142 fn int_tensor_handle(tensor: IntTensor<Self>) -> Self::Handle {
143 tensor
144 }
145
146 fn bool_tensor_handle(tensor: BoolTensor<Self>) -> Self::Handle {
147 tensor
148 }
149
150 fn quantized_tensor_handle(tensor: QuantizedTensor<Self>) -> Self::Handle {
151 tensor
152 }
153}