Hi, what I've done so far is:
--buy a new hardisk (a 7200 rpm sata), used only for the data to be transmitted,, and XFS formatted ---check my PC's USB bus via gnuradio benchmark_usb.py, which says everything is ok up to 32MB/s (output follows) ---use short int interleaved I/Q instead of gr_complex but still a 342958080 bytes long file, meant to last 10.717 seconds if played back at 32 MB/s (8 complex Msps) lasts 13+ seconds while, instead it should be: 342958080 / 32e6 = 10.717 seconds samples are damn slow to come out..this is quite frustrating as my application needs a true 8 MHz bandwidth... scaling down to 4 MHz is instead fine.. as I really get the right 342958080 / 16e6 = 21.435 seconds what else should I check/modify? any hint? the file meant to last 10.717 secs if played back with usrp_interp = 16 is at http://wwvince.interfree.it/of.dump please, if anybody manages to send it out in not more than 11 seconds, le me know.. it would be really good news... thanks vincenzo BENCHMARK_USB OUTPUT: [EMAIL PROTECTED] gr-mystuff]# ./monitor.sh Testing 2MB/sec... usb_throughput = 2M ntotal = 1000000 nright = 986311 runlength = 947654 delta = 52346 OK Testing 4MB/sec... usb_throughput = 4M ntotal = 2000000 nright = 1997961 runlength = 1997961 delta = 2039 OK Testing 8MB/sec... usb_throughput = 8M ntotal = 4000000 nright = 3997979 runlength = 3997979 delta = 2021 OK Testing 16MB/sec... usb_throughput = 16M ntotal = 8000000 nright = 7997477 runlength = 7997477 delta = 2523 OK Testing 32MB/sec... usb_throughput = 32M ntotal = 16000000 nright = 15995879 runlength = 15991525 delta = 8475 OK Max USB/USRP throughput = 32MB/sec -- Vincenzo Pellegrini
_______________________________________________ Discuss-gnuradio mailing list Discuss-gnuradio@gnu.org http://lists.gnu.org/mailman/listinfo/discuss-gnuradio
