Zixuan Jiang
Zixuan Jiang
## Generality * The 2d implementation should also support 1d input. * What if input is not a power of 2 ## Specificity * Optimization for cases where `M =...
* multiplying scale in `precomputeExpk` * zero paddings to avoid branch divergence * in-place or out-of-place cufft, especially in idct * number of threads in idct `M/2 * N/2` or...