[cfe-dev] RFC: First-class Matrix type

Andrea Bocci via cfe-dev cfe-dev at lists.llvm.org
Thu Oct 11 16:09:55 PDT 2018


On Fri, 12 Oct 2018, 00:44 Richard Smith via cfe-dev, <
cfe-dev at lists.llvm.org> wrote:

>
> On the LLVM IR side, I'm personally unconvinced that we should model
> matrices in the IR directly as a new first-class type, unless there's some
> target out there that has matrix operations in hardware / matrix registers,
> but IR is not really my area of expertise so give that opinion as much or
> little weight as you see fit.
>

Recent NVIDIA GPUs do support some matrix operations natively in the
hardware: https://devblogs.nvidia.com/programming-tensor-cores-cuda-9/ .

Ciao,
.Andrea

>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/cfe-dev/attachments/20181012/3a9ce6c1/attachment-0001.html>


More information about the cfe-dev mailing list