4 ms·
In this context I would imagine it's constructing source code for a kernel- the engine that implements a step in a neural network- that is closer to optimal. S
by dekhn 3y ago
In this context I would imagine it's constructing source code for a kernel- the engine that implements a step in a neural network- that is closer to optimal. See https://cuda.juliagpu.org/stable/tutorials/performance/ https://cuda.juliagpu.org/stable/tutorials/performance/ for related work
- Zentrik 3y agohttps://cuda.juliagpu.org/stable/tutorials/performance/ https://cuda.juliagpu.org/stable/tutorials/performance/ provides various tips that someone who has written a kernel can use to speed it up, like using 32 bit integers and minimising runtime exceptions. Perhaps, I'm misunderstanding but it's not part of direct kernel generation, whatever that is.