Enum Concurrency 

pub enum Concurrency {
    Cpu,
    Gpu,
    GpuPlusCpu,
}

Backend selection for Rationer similarity work.

Picked at construction; can’t change for a given handle (the resource set is wired up at build() time). Re-create the handle if you need a different concurrency mode.

Variants

Cpu

Pure CPU — single-threaded if threads(1), multi-threaded rayon otherwise. Works on every platform; the only mode available with feature = "gpu" disabled.

Gpu

Allow the Metal GPU where it measured a net win. In practice that is only cluster_canonicals on a single group above DFGPU_CLUSTER_THRESHOLD (~1.1–1.4× CPU, peaking on mid-size canonical-Python corpora). ratio_many and cluster_canonicals_multi stay on CPU under this mode too — the GPU matching_stats offload lost on both at every size benchmarked (their GPU paths are env-opt-in only). On non-Metal platforms: Cpu.

GpuPlusCpu

Default. Same routing as Gpu (GPU only for big single-group cluster_canonicals, everything else CPU), plus rayon-parallel CPU for the per-pair longest_in recursion on the GPU output. Behaves identically to Gpu for the currently GPU-enabled path; kept as a distinct mode so heterogeneous overlap can be re-enabled per-path via the env thresholds without an API change.

Trait Implementations

impl Clone for Concurrency

fn clone(&self) -> Concurrency

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for Concurrency

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

Formats the value using the given formatter.

impl Default for Concurrency

fn default() -> Self

Returns the “default value” for a type.

impl PartialEq for Concurrency

fn eq(&self, other: &Concurrency) -> 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 Concurrency

impl Eq for Concurrency

impl StructuralPartialEq for Concurrency