[LAU] jdelay - measured latency of a usb interface

[James Cameron]

On Tue, Mar 24, 2009 at 11:39:03PM -0700, Ken Restivo wrote:
> Is it possible that the USB protocol sets a *minimum* on latency,
> and thus any speed improvements in the USB hardware are essentially
> moot, limited by this lower bound imposed by the USB protocol?

Yes, there is a lower bound.  It depends on the protocol version being
used.

Also, don't blindly believe the latency values displayed, as these may
be estimates.  If you have the means, measure the latency using external
analysis.

--

Diversion.

We've been reading about USB latency in the past couple of days on the
One Laptop Per Child project, where USB to Ethernet devices are attached
to tiny school servers.  If you'd like to follow the technical details
...

Problem statement:
http://lists.laptop.org/pipermail/devel/2009-March/023887.html

Solution:
http://lists.laptop.org/pipermail/server-devel/2009-March/003105.html
(it was a USB 1.1 device)

And for the truly interested, the wireless chipset is also on USB, and
the Linux kernel appears to do half the possible data rate, given that
the OpenFirmware can achieve so much more with full control of the
hardware:

http://lists.laptop.org/pipermail/devel/2009-March/023899.html

(the data rate that is achieved by the kernel is well within the needs
of the target market though).

-- 
James Cameron    mailto:quozl at us.netrek.org     http://quozl.netrek.org/