The issue of CPU overhead for managing data transfers goes way back. I was arguing the superior transfer rate of SCSI over ATA years ago. Fortunately, I could demonstrate the difference side-by-side in our test configuration. Not everything in tech is as simple as some imagine it. The extra latency from those old x86 processors made a huge difference, making ATA clearly slower.
Thank you, Mark Snyder -----Original Message----- The transfer rate for the latest FireWire and eSATA versions don't matter much because both are high enough to not be creating the data- transfer bottleneck. That is why faster versions of these standards have been proposed BUT NOT IMPLEMENTED. It would be a waste of effort to do so. If anybody bothered to look at the charts on tomshardware that I linked to they would have seen both FireWire and eSATA drives in mixed order at the top of the charts. This amply demonstrates that FireWire vs eSATA does not matter. Other parts of the data channel are what is limiting the data rate. One example. Both FireWire and SCSI control the data flow in hardware. ATA and eSATA have the CPU managing the data transfer. So FireWire and SCSI can maintain a high data rate irrespective of what is happening in the CPU and ATA and eSATA can't. That may account for the inconsistent eSATA results that were observed. ************************************************************************ * ** List info, subscription management, list rules, archives, privacy ** ** policy, calmness, a member map, and more at http://www.cguys.org/ ** ************************************************************************ * ************************************************************************* ** List info, subscription management, list rules, archives, privacy ** ** policy, calmness, a member map, and more at http://www.cguys.org/ ** *************************************************************************
