README.md

cuFFT — Fast Fourier Transform Examples

4 examples covering 1D, 2D, and 3D FFTs with both complex and real data. Enable with -Dcufft=true.

Build & Run

zig build run-cufft-<name> -Dcufft=true

zig build run-cufft-fft_1d_c2c -Dcufft=true
zig build run-cufft-fft_1d_r2c -Dcufft=true
zig build run-cufft-fft_2d    -Dcufft=true
zig build run-cufft-fft_3d    -Dcufft=true

---

Examples

Example	File	Transform	Description
fft_1d_c2c	fft_1d_c2c.zig	C2C	1D complex-to-complex FFT + inverse round-trip
fft_1d_r2c	fft_1d_r2c.zig	R2C	1D real-to-complex FFT with frequency-domain filtering
fft_2d	fft_2d.zig	C2C	2D complex FFT on a matrix
fft_3d	fft_3d.zig	C2C	3D complex FFT on a volume

---

Key API

const cufft = @import("zcuda").cufft;

// 1D complex-to-complex
const plan = try cufft.CufftPlan.plan1d(n, .c2c, 1);
defer plan.deinit();

try plan.execC2C(d_input, d_output, .forward);
try plan.execC2C(d_output, d_input, .inverse); // round-trip

// 1D real-to-complex
const plan_r2c = try cufft.CufftPlan.plan1d(n, .r2c, 1);
try plan_r2c.execR2C(d_real, d_complex);

// 2D
const plan_2d = try cufft.CufftPlan.plan2d(rows, cols, .c2c);
try plan_2d.execC2C(d_in, d_out, .forward);

Transform Types

Type	Input → Output	Use Case
c2c	Complex → Complex	General FFT / IFFT
r2c	Real → Complex	Real-valued signals (saves memory)
c2r	Complex → Real	Inverse of r2c
z2z	Double complex → Double complex	Double-precision FFT

→ Full API reference: docs/cufft/README.md