Struct CudaDeviceInfo 

pub struct CudaDeviceInfo {
    pub ordinal: u32,
    pub name: String,
    pub total_mem_bytes: u64,
    pub compute_capability: (u32, u32),
    pub multiprocessor_count: u32,
    pub max_threads_per_block: u32,
    pub warp_size: u32,
    pub supports_unified_memory: bool,
    pub supports_f64: bool,
    pub driver_version: String,
}

Information about the selected CUDA device.

Fields

ordinal: u32

CUDA device ordinal (0-indexed).

name: String

Device name from cuDeviceGetName.

total_mem_bytes: u64

Total global memory in bytes (cuDeviceTotalMem).

compute_capability: (u32, u32)

Compute capability as (major, minor).

multiprocessor_count: u32

Number of CUDA streaming multiprocessors.

max_threads_per_block: u32

Maximum threads per block.

warp_size: u32

Warp size (always 32 on current NVIDIA hardware).

supports_unified_memory: bool

Whether the device supports unified memory (Compute Capability ≥ 3.0).

supports_f64: bool

Whether the device supports FP64 (cuDeviceGetAttribute CUDA_DEVICE_ATTRIBUTE_DOUBLE).

driver_version: String

CUDA driver version string.

Trait Implementations

impl Clone for CudaDeviceInfo

fn clone(&self) -> CudaDeviceInfo

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for CudaDeviceInfo

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

Formats the value using the given formatter.

impl Default for CudaDeviceInfo

fn default() -> CudaDeviceInfo

Returns the “default value” for a type.