The following table describes the latency for the AXI DataMover
core. The table provides performance information for a typical configuration.
Throughput testing consisted of eight parent commands loaded into the AXI DataMover core with each command having a BTT value as 1 MB and
each channel operating simultaneously (full duplex). The core was configured for
synchronous operation meaning m_axis_mm2s_cmdsts_aclk = m_axis_s2mm_cmdsts_awclk = m_axi_mm2s_aclk = m_axi_s2mm_aclk.
The core configuration used to generate the throughput data is as follows:
- Memory Map Data Width = 32
- Streaming Data Width = 32
- No unaligned access
- Store and Forward enabled
| AXI DataMover Channel | Primary Clock Frequency | Packet Size | Maximum Total Data Throughput (MB/sec) | Percent of Theoretical |
|---|---|---|---|---|
| MM2S | 100 | 1 MB | 391.27 | 97.75% |
| S2MM | 100 | 1MB | 391.27 | 97.75% |