3 ms·

The paper is about integer multiplication, not float

• oofbey 2y ago
  Why would that matter? I understood the point was to speed up matrix multiplication by doing the adds and multiplies in a different order. Shouldn't matter whether the datatype is int, float, complex, whatever.
  • sifar 2y ago
    One can use this techniques to optimize the multiplier inside the FP FMA unit. However this cannot be used to multiply two floating point numbers as FP arithmetic is not associative.
    • saagarjha 2y ago
      I'm pretty sure a lot of people doing matrix multiplications are ok with non-associative floating point arithmetic?