Re: [Discuss-gnuradio] Alternative interface from PC to a USRP.

[Matt Ettus]

> The USRP is capable of sampling 256MS/s (64MS/s across four channels), so you 
> need 256MB/s to not lose data, if you are interested in, say, doing a 
> continuous spectrum capture across a 128MHz band.  We have an application for 
> Radio Astronomy that involves interferometry where bandwidth is king; the 
> researcher would love to get two polarizations times two frequencies times 
> two antennas (8 channels) with 200MHz of bandwidth per channel; this is a lot 
> of data, yes, but he claims that you can't have too much bandwidth doing what 
> he is doing.  For that application, we have put in a grant proposal for an 
> SRC MAPstation with 16 8-bit 200Ms/s ADC's.  The SRC MAPstation uses three 
> high-speed (200MHz) Xilinx FPGA's to do massively parallel calculations; the 
> interface to the PC is SRC's proprietary SNAP interconnect that plugs into a 
> DDR-SDRAM socket on the PC motherboard.  They sell 'servers' with up to 128 
> MAP processors and crossbar switching of the SNAP interconnect for hundreds 
> of gigaflops of performance and up to 6GB/s I/O bandwidth for things like 
> ADC's.  They are not cheap.


Sounds like a very interesting application.  However, getting all those
bits in and out of the computer would only solve half the problem.  You
need to be able to perform computations on that data, and your PC would
not be able to keep up.  The best solution is to preprocess in the FPGA,
which you are doing with the MAPstation.  What is the data rate coming
out of the MAPstation and into the computer?

Matt