Hi, I was wandering about the status of the llvm.experimental.vector.reduce.* intrinsics. Are all back-ends supporting those intrinsics or are they still in a very "experimental" state? Thanks, Michael