As explained previously, the paging latency for FFT with input size equal to 24 MB is 61.279 sec, or 11.24 ms per page transfer. From these, 1.6 ms were spent during protocol processing and 9.64 ms were spent transferring each page on the Ethernet.
Previous measurements have reported that a 4 KByte page takes about 45 ms over an Ethernet for each pagein . Of those 45 ms, 19 ms were spent on TCP overhead, 4 ms were spent on Mach IPC overhead, 7.2 ms were spend on the Ethernet, and the rest were spent on the computer's I/O bus. The total software latency of our implementation, is only 1.6 ms. The reason for this significant difference in performance is threefold:
In general, although our approach may have less flexibility than a full-fledged user-level pager, it has much better performance. Moreover, our device-driver implementation provides better performance than traditional (local) disk paging, while user-level implementations have not reported performance results to support similar claims .