search

xtensor-stack / xtensor

C++ tensors with broadcasting and lazy computing
BSD 3-Clause "New" or "Revised" License
3.38k stars 401 forks source link

Use simd_steppers in accumulator and reducers #960

Open wolfv opened 6 years ago

wolfv commented 6 years ago

One idea to make the step_leading more generic is to rely on the static layout of the expression to either increment by 1 or use the (real) leading stride (first or last of strides() depending on dynamic layout).