nn_mm_f16()

void nn_mm_f16	(	Tensor2D_F16 *	y,
		const Tensor2D_F16 *	x1,
		const Tensor2D_F16 *	x2
	)