Hi, the discussion went on and the patch to add LLVM matrix intrinsics got committed: https://reviews.llvm.org/rG526244b187d23351056de6aba715dc3d44668b1c <https://reviews.llvm.org/rG526244b187d23351056de6aba715dc3d44668b1c>
This discussion about adding matrix support to the clang frontend explicitly references Eigen (and benchmarks against it), so though I'm going to share it: http://lists.llvm.org/pipermail/cfe-dev/2020-January/064371.html <http://lists.llvm.org/pipermail/cfe-dev/2020-January/064371.html> Cheers, David > On 20. Nov 2019, at 22:12, Gael Guennebaud <[email protected]> wrote: > > > Thanks for sharing, it will be interesting to follow where this is going. > > Gaël. > > On Wed, Nov 20, 2019 at 8:38 PM David Tellenbach > <[email protected] <mailto:[email protected]>> > wrote: > Hi all, > > just wanted to point you to an interesting proposal on the LLVM mailing list: > Dedicated LLVM intrinsics for matrix math > (http://lists.llvm.org/pipermail/llvm-dev/2019-October/136240.html > <http://lists.llvm.org/pipermail/llvm-dev/2019-October/136240.html>). > > On Tuesday a first patch got into review, so this is relatively concrete: > https://reviews.llvm.org/D70456 <https://reviews.llvm.org/D70456>. Eigen even > got mentioned here: https://reviews.llvm.org/D70456#inline-635182 > <https://reviews.llvm.org/D70456#inline-635182> > > Cheers, > David
