摘要 |
Methods and structures are provided for multi-lane data communication with measurable latency. In a particular embodiment, part of a parallel data set is transmitted from a first device to a second device on a plurality of slave lanes and another part of the data set is transmitted from the first device to the second device on a master lane. A known master delay is applied to data in the master lane that is greater than or equal to the known maximum skew between the lanes. The slave lanes are delayed as needed to align their data with the master lane. In one embodiment, part of the known master lane delay is applied on the first device and another part is applied on the second device. In another embodiment, all of the known master lane delay is applied on the first device and none of it is applied on the second device. In another embodiment, all of the known master lane delay is applied on the second device and none of it is applied on the first device. In another embodiment, part or all of the master lane delay is applied on the link between devices.
|