Trait numa_gpu::runtime::hw_info::CudaDeviceInfo

pub trait CudaDeviceInfo {
    fn sm_cores(&self) -> Result<u32>;
    fn cores(&self) -> Result<u32>;
    fn concurrent_managed_access(&self) -> Result<bool>;
    fn clock_rate(&self) -> Result<u32>;
    fn memory_clock_rate(&self) -> Result<u32>;
}

Extends Rustacuda’s Device with methods that provide additional hardware information.

Required methods

fn sm_cores(&self) -> Result<u32>

Returns the number of cores per streaming multiprocessor

fn cores(&self) -> Result<u32>

Returns the total number of cores that are in the device

fn concurrent_managed_access(&self) -> Result<bool>

Returns true if concurrent managed access is supported by the device

fn clock_rate(&self) -> Result<u32>

Returns the default clock rate of the streaming multiprocessor in megahertz

fn memory_clock_rate(&self) -> Result<u32>

Returns the default memory clock rate of the GPU in megahertz

Implementations on Foreign Types

impl CudaDeviceInfo for Device

fn sm_cores(&self) -> Result<u32>

fn cores(&self) -> Result<u32>

fn concurrent_managed_access(&self) -> Result<bool>

fn clock_rate(&self) -> Result<u32>

fn memory_clock_rate(&self) -> Result<u32>

Implementors