Struct faiss::gpu::StandardGpuResources[][src]

pub struct StandardGpuResources { /* fields omitted */ }
Expand description

Standard GPU resources descriptor.

Examples

GPU resources are meant to be passed to an index implementation’s into_gpu or to_gpu methods.

use faiss::{StandardGpuResources, MetricType};
use faiss::index::flat::FlatIndex;

let gpu = StandardGpuResources::new()?;
let index = FlatIndex::new(64, MetricType::L2)?;
let gpu_index = index.into_gpu(&gpu, 0)?;

Since GPU implementations are not thread-safe, attempting to use the GPU resources from another thread is not allowed.

use faiss::{GpuResources, StandardGpuResources};
use faiss::index::flat::FlatIndex;
use std::sync::Arc;
use std::thread;

let gpu = Arc::new(StandardGpuResources::new()?);
let gpu_rc = gpu.clone();
thread::spawn(move || {
    let index = FlatIndex::new_l2(64)?;
    let gpu_index = index.into_gpu(&*gpu_rc, 0)?; // will not compile
    Ok(())
});

Other than that, indexes can share the same GPU resources, so long as neither of them cross any thread boundaries.

use faiss::{GpuResources, StandardGpuResources, MetricType, index_factory};

let mut gpu = StandardGpuResources::new()?;
let index1 = index_factory(64, "Flat", MetricType::L2)?
    .into_gpu(&gpu, 0)?;
let index2 = index_factory(32, "Flat", MetricType::InnerProduct)?
    .into_gpu(&gpu, 0)?;

Implementations

Create a standard GPU resources object.

Trait Implementations

Formats the value using the given formatter. Read more

Obtain a raw pointer to the native GPU resources object.

Disable allocation of temporary memory; all temporary memory requests will call cudaMalloc / cudaFree at the point of use Read more

Specify that we wish to use a certain fixed size of memory on all devices as temporary memory Read more

Set amount of pinned memory to allocate, for async GPU <-> CPU transfers Read more

Obtain a raw pointer to the native GPU resources object.

Disable allocation of temporary memory; all temporary memory requests will call cudaMalloc / cudaFree at the point of use Read more

Specify that we wish to use a certain fixed size of memory on all devices as temporary memory Read more

Set amount of pinned memory to allocate, for async GPU <-> CPU transfers Read more

Obtain a raw pointer to the native GPU resource provider object.

Auto Trait Implementations

Blanket Implementations

Gets the TypeId of self. Read more

Immutably borrows from an owned value. Read more

Mutably borrows from an owned value. Read more

Performs the conversion.

Performs the conversion.

The type returned in the event of a conversion error.

Performs the conversion.

The type returned in the event of a conversion error.

Performs the conversion.