I notice on the Wiki that the conjecture about the bandwidth of
the USB2 has changed to an assurtion (vis-a-vis hdtv.) That was
neat to see, but it left me with a question (which I hate to spam
the Wiki with.) Is it the case that the USB2 bandwidth is
sufficient for 1/1, but the overhead of related CPU-side
processing brings the ratio down?
That is correct. The USRP is fast enough to handle HDTV in real time, but the
PC is not (yet). I suspect that with a little bit of tuning, and some offload
to the USRP FPGA, that it could be done.
BTW, Eric and I are in Utah right now demonstrating the USRP for an NSF crowd.
Matt