"Core ML is optimized for on-device performance, which minimizes memory footprint and power consumption." This is major, if they have managed to achieve it rea

"Core ML is optimized for on-device performance, which minimizes memory footprint and power consumption." This is major, if they have managed to achieve it reasonably. But before opening a Sekt, I want to see some benchmarks. :)