Obviously there is overhead so you'll never see all of that but I would have thought it would be better than 50-60% of the theoretical maximum.
Right, if that was the case that would mean it was terribly badly designed... oh.
There are limits in the USB2 standard on how often you can service any one endpoint, even if it's the only active one. Someone who was on some of the committees told me that this wasn't what they
meant, but it's what they
wrote, and so that's what got implemented. USB3 is better in this regard, not just because the overall bandwidth is much higher, but because the channel-occupancy rules are less silly.
Peter
Previous Topic
Index
