Struct EpistemicGpuMaterializationTrace 

pub struct EpistemicGpuMaterializationTrace {
    pub materialized_candidates: usize,
    pub world_view_slots_written: usize,
    pub kernel_launches: u32,
    pub host_write_ops: u32,
    pub kernel_timing: EpistemicGpuKernelTimingTrace,
}

Trace proving accepted-candidate materialization staging used a GPU kernel.

Fields

materialized_candidates: usize

Number of candidate rows materialized on device.

world_view_slots_written: usize

World-view slots written by the kernel.

kernel_launches: u32

Materialization kernel launches.

host_write_ops: u32

Host writes used by materialization. Accepted GPU execution requires zero.

kernel_timing: EpistemicGpuKernelTimingTrace

CUDA-event timing for the launched kernel.

Implementations

impl EpistemicGpuMaterializationTrace

pub fn for_count(candidate_count: usize) -> Result<Self>

Build a materialization trace for a bounded device launch.

pub const fn with_kernel_timing( self, kernel_timing: EpistemicGpuKernelTimingTrace, ) -> Self

Attach CUDA-event timing captured by the runtime launch path.

Trait Implementations

impl Clone for EpistemicGpuMaterializationTrace

fn clone(&self) -> EpistemicGpuMaterializationTrace

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for EpistemicGpuMaterializationTrace

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

Formats the value using the given formatter.

impl PartialEq for EpistemicGpuMaterializationTrace

fn eq(&self, other: &EpistemicGpuMaterializationTrace) -> bool

Tests for self and other values to be equal, and is used by ==.

fn ne(&self, other: &Rhs) -> bool

Tests for !=. The default implementation is almost always sufficient, and should not be overridden without very good reason.

impl Copy for EpistemicGpuMaterializationTrace

impl Eq for EpistemicGpuMaterializationTrace

impl StructuralPartialEq for EpistemicGpuMaterializationTrace