Re: [Starpu-devel] Support for half precision

[Hatem Ltaief <hatem.ltaief@kaust.edu.sa>]

We would like to use a tile matrix in single precision to host tiles in 
single precision as well as tiles in half precision. When about to operate on 
tiles containing half precision data, then only half of the data tile is 
meaningful, we would like starpu to transfer only the meaningful data from 
the tiles when needed, ie full data tile for single precision and half data 
tile for half precision.
Basically, you are saying that we can allocate our overall matrix with tile 
data structure in single precision, and at the time of the starpu data 
registration done on a tile-basis, we should specify the proper element size.
Is that correct?

Sent from my iPhone

> On Oct 19, 2018, at 18:08, Samuel Thibault <samuel.thibault@inria.fr> wrote:
> 
> Hatem Ltaief, le ven. 19 oct. 2018 13:27:57 +0000, a ecrit:
>> There is an explicit cuda support for half precision (16bits) to run on 
>> Volta V100 GPU with tensor cores.
>> How should we proceed with StarPU to exploit this data type, as far as the 
>> data transfer between GPUs is concerned?
> 
> You just need to fix the element size in matrix registration etc.
> 
> Samuel