No comparison of speed?
Andrew Carlson
1

Both work on CUDA and CUDNN, so the speed should be close for both on GPUs.

One clap, two clap, three clap, forty?

By clapping more or less, you can signal to us which stories really stand out.