yes that’s what i am working on these days but there is a need for a generally available neural chip (see google’s coral as one attempt). in my tests, esp32s3 is very very slow for any model with conv2d involved.
i just want a tiiiny gpu for $10 so i can run smaller models at higher speed than possible with xtensa/rp2040 having limited simd support etc.
i just want a tiiiny gpu for $10 so i can run smaller models at higher speed than possible with xtensa/rp2040 having limited simd support etc.