A company is deploying a multi-instance application within AWS that requires minimal latency between the instances. What should a solutions architect recommend?