pub struct FftPlannerAvx<T: FftNum> { /* private fields */ }
Expand description
The AVX FFT planner creates new FFT algorithm instances which take advantage of the AVX instruction set.
Creating an instance of FftPlannerAvx
requires the avx
and fma
instructions to be available on the current machine, and it requires RustFFT’s
avx
feature flag to be set. A few algorithms will use avx2
if it’s available, but it isn’t required.
For the time being, AVX acceleration is black box, and AVX accelerated algorithms are not available without a planner. This may change in the future.
// Perform a forward Fft of size 1234, accelerated by AVX
use std::sync::Arc;
use rustfft::{FftPlannerAvx, num_complex::Complex};
// If FftPlannerAvx::new() returns Ok(), we'll know AVX algorithms are available
// on this machine, and that RustFFT was compiled with the `avx` feature flag
if let Ok(mut planner) = FftPlannerAvx::new() {
let fft = planner.plan_fft_forward(1234);
let mut buffer = vec![Complex{ re: 0.0f32, im: 0.0f32 }; 1234];
fft.process(&mut buffer);
// The FFT instance returned by the planner has the type `Arc<dyn Fft<T>>`,
// where T is the numeric type, ie f32 or f64, so it's cheap to clone
let fft_clone = Arc::clone(&fft);
}
If you plan on creating multiple FFT instances, it is recommended to re-use the same planner for all of them. This is because the planner re-uses internal data across FFT instances wherever possible, saving memory and reducing setup time. (FFT instances created with one planner will never re-use data and buffers with FFT instances created by a different planner)
Each FFT instance owns Arc
s to its internal data, rather than borrowing it from the planner, so it’s perfectly
safe to drop the planner after creating Fft instances.
Implementations§
Source§impl<T: FftNum> FftPlannerAvx<T>
impl<T: FftNum> FftPlannerAvx<T>
Sourcepub fn new() -> Result<Self, ()>
pub fn new() -> Result<Self, ()>
Constructs a new FftPlannerAvx
instance.
Returns Ok(planner_instance)
if we’re compiling for X86_64, AVX support was enabled in feature flags, and the current CPU supports the avx
and fma
CPU features.
Returns Err(())
if AVX support is not available.
Sourcepub fn plan_fft(
&mut self,
len: usize,
direction: FftDirection,
) -> Arc<dyn Fft<T>>
pub fn plan_fft( &mut self, len: usize, direction: FftDirection, ) -> Arc<dyn Fft<T>>
Returns a Fft
instance which uses AVX instructions to compute FFTs of size len
.
If the provided direction
is FftDirection::Forward
, the returned instance will compute forward FFTs. If it’s FftDirection::Inverse
, it will compute inverse FFTs.
If this is called multiple times, the planner will attempt to re-use internal data between calls, reducing memory usage and FFT initialization time.
Sourcepub fn plan_fft_forward(&mut self, len: usize) -> Arc<dyn Fft<T>>
pub fn plan_fft_forward(&mut self, len: usize) -> Arc<dyn Fft<T>>
Returns a Fft
instance which uses AVX instructions to compute forward FFTs of size len
.
If this is called multiple times, the planner will attempt to re-use internal data between calls, reducing memory usage and FFT initialization time.
Sourcepub fn plan_fft_inverse(&mut self, len: usize) -> Arc<dyn Fft<T>>
pub fn plan_fft_inverse(&mut self, len: usize) -> Arc<dyn Fft<T>>
Returns a Fft
instance which uses AVX instructions to compute inverse FFTs of size `len.
If this is called multiple times, the planner will attempt to re-use internal data between calls, reducing memory usage and FFT initialization time.