"iBizTalk" <ibiztalk@gmail.com> wrote in message
news:4FF85DFD-6796-4CD1-A1CB-CDCA739FD4A3@microsoft.com...
> Hello,
>
> Does anyone out there know how I can test the speed of my USB ports
> without buying hardware? Is there any software out there to do this
> (hopefully free)?
>
Are you perhaps worried that you are not moving data around at the claimed
480 Mbps. If so, you would be right. 480 Mbps is the bit data rate of all
the communication that occurs over the USB bus. Out of that has to come a
certain amount of house keeping - this is the USB host controller actually
controlling all the hubs and peripheral devices, deciding what can talk and
when. Of greater impact is that USB is implemented as a single bus, thus
communication can only take place in one direction at any one time. The 480
Mbps therefore represents the total data rate in *both* directions [1]. The
remaining significant bottleneck is that the USB host port requires
considerable processor support and if the processor isn't ready to service
it, then it has to wait.
Contrast Firewire, where although it is only rated at 400 Mbps (in the
common IEE1394a incarnation) the actual practical data rate is significantly
higher because, being a dual bus system, you get 400 Mbps in *each*
direction. Equally significantly the Firewire ports are able to operate
more or less autonomously without processor support and thus don't have to
wait. The housekeeping requirements are largely similar buthere is no need
to hand over data direction. In practice, the differences are usage
dependant, but a 33% increase in speed is the sort of minimum that you can
expect from Firewire.
[1] USB1 had another drawback in that communication with slow peripherals
(1.5 Mbps) operated at that speed from end to end which had a major impact
on the faster (12 Mbps) communications. USB2 solved this problem by
requiring USB2 hubs to buffer the data and then up convert it to 480 Mbps
for transmission to the host, and similarly for reception from the host.
