Struct WgpuRuntime

pub struct WgpuRuntime;

Runtime that uses the wgpu crate with the wgsl compiler. This is used in the Wgpu backend. For advanced configuration, use init_setup to pass in runtime options or to select a specific graphics API.

Trait Implementations

impl Debug for WgpuRuntime

fn fmt(&self, f: &mut Formatter<'_>) -> Result<(), Error>

Formats the value using the given formatter.

impl Runtime for WgpuRuntime

type Compiler = AutoCompiler

The compiler used to compile the inner representation into tokens.

type Server = WgpuServer

The compute server used to run kernels and perform autotuning.

type Channel = MutexComputeChannel<WgpuServer>

The channel used to communicate with the compute server.

type Device = WgpuDevice

The device used to retrieve the compute client.

fn client( device: &<WgpuRuntime as Runtime>::Device, ) -> ComputeClient<<WgpuRuntime as Runtime>::Server, <WgpuRuntime as Runtime>::Channel>

Retrieve the compute client from the runtime device.

fn name( client: &ComputeClient<<WgpuRuntime as Runtime>::Server, <WgpuRuntime as Runtime>::Channel>, ) -> &'static str

The runtime name on the given device.

fn supported_line_sizes() -> &'static [u8]

Returns the supported line sizes for the current runtime’s compiler.

fn max_cube_count() -> (u32, u32, u32)

Returns the maximum cube count on each dimension that can be launched.

fn device_id(device: &<WgpuRuntime as Runtime>::Device) -> DeviceId

Fetch the id for the given device.

fn can_read_tensor(shape: &[usize], strides: &[usize]) -> bool

fn require_array_lengths() -> bool

Return true if global input array lengths should be added to kernel info.

fn line_size_elem(elem: &Elem) -> impl Iterator<Item = u8> + Clone

Returns all line sizes that are useful to perform IO operation on the given element.