[USRP-users] Bound in streaming X310 through PCIE

Raj Bhattacharjea raj.b at gatech.edu
Wed Mar 9 15:49:43 EST 2016


Marcus,

I had a suspicion that uhd_fft and osmocom_fft weren't actually crunching
FFTs on all that data, thanks for confirming it. I agree, it really depends
on what you're doing.

Here's another data point for reference: using PCIe to a single X310, I am
able to actually FFT every sample using gr-fosphor @ 66.666667 MSamp/s on
an Intel i7-3770, which is a desktop class CPU that is three generations
old. I'm using Intel's OpenCL CPU implementation for RHEL and SLES, but
shoehorned into working on Ubuntu 14.04:

https://bitbucket.org/snippets/rajb245/epGzn

I am about to go try this on a more powerful Xeon setup with 10 GigE to see
what is actually possible.

-- 
Raj Bhattacharjea, PhD
Georgia Tech Research Institute
Information and Communications Laboratory
http://www.prism.gatech.edu/~rb288/
404.407.6622
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.ettus.com/pipermail/usrp-users_lists.ettus.com/attachments/20160309/0c9f7e97/attachment-0002.html>


More information about the USRP-users mailing list