On 09/05/2018 11:19 AM, Martin Braun via USRP-users wrote:
If I may add some comments: We *do* have converters from sc12 to sc16,
so I'm not sure what's up with that. If you have a modern computer, 56
Msps isn't considered a massive amount (GNU Radio, for example, should
have no issues chomping away that much on a modern PC).
I suspect you might have stumbled across some corner-case bug, although
it's interesting that we haven't seen this more often.
For certain values of "modern", and again I'll point out
*depending on what you're doing with the samples*.
It's one thing to chew on a large amount of bandwidth doing a "stuttered"
FFT implementation (which is what GR does for the FFT sink, BTW).
It's quite another to be paying attention and "doing things" to every
last sample
at the input sample rate.
There can be differences of one or two orders of magnitude depending on
exactly what
you're doing with the samples.
Doing writes to the file-system at high input rates is another issue
where CPU beefiness is only
part of the picture--your drives and PCI subsystem have to be up to
snuff as well.
I'll use another of my famous analogies, because I just love them so much :)
You buy a car capable of doing 300km/h without any issue. But ignoring
legal issues for the moment, you don't
seem to be able to find any streets for which you're able to achieve
that without having a fiery crash...
_______________________________________________
USRP-users mailing list
USRP-users@lists.ettus.com
http://lists.ettus.com/mailman/listinfo/usrp-users_lists.ettus.com
