Closed johnplatts closed 3 months ago
Reimplemented I32/U32 SumOfMulQuadAccumulate on NEON_BF16 since the vdot[q]_s32 and vdot[q]_u32 intrinsics are guaranteed to be available on the NEON_BF16 target, which requires support for the AArch64 DotProd extension.
Reimplemented I32/U32 SumOfMulQuadAccumulate on NEON_BF16 since the vdot[q]_s32 and vdot[q]_u32 intrinsics are guaranteed to be available on the NEON_BF16 target, which requires support for the AArch64 DotProd extension.