CUFFT in parallel

Accelerated Computing GPU-Accelerated Libraries
1

Hello dear NVIDIA community,

I am implementing a code with CUFFT library,

setting the plan as:
#define BATCH 2
#define FFT_size 512
cufftPlan1d(&plan, FFT_size, CUFFT_C2C, BATCH);
cufftExecC2C(plan, d_signal_in, d_signal_out, CUFFT_FORWARD);

My questions are:

  1. How many GPU threads, blocks and dims are involved?
  2. Is it possible to run such several operations simultaneously e.g. call cufftExecC2C() 10 times or more ?

Thank you