nuvec-gpu 0.6.1

Metal runtime for nuvec: device, pipelines, buffers, command encoding, allocator and profiler
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

[package]
name = "nuvec-gpu"
description = "Metal runtime for nuvec: device, pipelines, buffers, command encoding, allocator and profiler"
exclude.workspace = true
version.workspace = true
edition.workspace = true
rust-version.workspace = true
license.workspace = true
repository.workspace = true
authors.workspace = true
homepage.workspace = true
readme.workspace = true
keywords.workspace = true
categories.workspace = true

[dependencies]
nuvec-kernels.workspace = true

[features]
# The experimental CUDA runtime (Linux): NVIDIA's cuda-rust, which needs CUDA ≥ 13.0 headers
# and libclang to build. Without it, or off Linux, `nuvec_gpu::cuda` does not exist.
cuda = ["nuvec-kernels/cuda", "dep:cuda-core", "dep:cuda-async", "dep:cutile", "dep:half"]

[target.'cfg(target_os = "linux")'.dependencies]
cuda-core = { version = "=0.4.0", optional = true }
cuda-async = { version = "=0.4.0", optional = true }
cutile = { version = "=0.4.0", optional = true }
half = { workspace = true, optional = true }

[target.'cfg(target_os = "macos")'.dependencies]
block2.workspace = true
libc.workspace = true
dispatch2.workspace = true
objc2.workspace = true
objc2-foundation.workspace = true
objc2-metal.workspace = true

# Document the Metal API, which only exists on Apple targets.
[package.metadata.docs.rs]
targets = ["aarch64-apple-darwin"]

[[example]]
name = "cuda_bench"
required-features = ["cuda"]

[lints]
workspace = true