Hi Jason, Adding FIFOs adds buffering which helps with any transient changes in throughput, such as over the 10 GigE connection. It gives the flow control more room to work with before an overflow occurs (on receive). On the transmit side the DMA FIFO usually fills that role.
On Fri, Sep 29, 2017 at 4:12 AM, Jason Matusiak via USRP-users < usrp-users@lists.ettus.com> wrote: > OK, thanks Martin. I guess I am still confused though (no surprise > there). You say that they're "generally useful." In what way? If I have > a flow graph that has 3 RFNoC blocks, what is the benefit of adding a FIFO > or two to it? > > Thanks! > > > > On 09/28/2017 03:10 PM, Martin Braun via USRP-users wrote: > > On Wed, Sep 27, 2017 at 01:42:22PM -0400, Jason Matusiak via USRP-users wrote: > > OK, dumb question, but I just can't come up with a good answer. I > understand that the RFNoC FIFOs are a must if you only have one NoC block > that you want to use and are using the GNURadio host [1]. So why do pretty > much most of the examples ALWAYS have at least one, and why would I want to > fill my bitfile with FIFOs, why not just use one? > > Note that this is a gr-ettus limitation, not RFNoC. And also, you only > need FIFOs if you have only a single block (PC -> Block -> PC). > > And you don't need to fill up with FIFOs, they're just generally useful > and thus we have that option. > > -- M > > > > > _______________________________________________ > USRP-users mailing > listUSRP-users@lists.ettus.comhttp://lists.ettus.com/mailman/listinfo/usrp-users_lists.ettus.com > > > > _______________________________________________ > USRP-users mailing list > USRP-users@lists.ettus.com > http://lists.ettus.com/mailman/listinfo/usrp-users_lists.ettus.com > >
_______________________________________________ USRP-users mailing list USRP-users@lists.ettus.com http://lists.ettus.com/mailman/listinfo/usrp-users_lists.ettus.com
