[USRP-users] Success report: B200/USB 3.0 card/Core i7-2600

Garver, Paul W garverp at gatech.edu
Wed Oct 1 12:35:30 EDT 2014


In my experience 10sec is not long enough to test if the drive can keep up for longer recording sessions (>1min).  I observe (via iostat) the kernel buffering up the data but it won't actually write it out until a few seconds later.  If you keep streaming data, you can never write fast enough to get the average to the required MB/sec (100MBytes/sec @ 25MSPS)  I'd be curious to know if a 10 minute test results in a perfect output on your system.  I ended up modifying rx_samples_to_file to call sync_file_range() which has performed near flawlessly at 25MSPS writing ~1TB of samples to a Samsung 840 Evo mSATA SSD at the cost of CPU usage.  I'd be curious to hear other's experiences for long records on the B200/210.

PWG

Date: Tue, 30 Sep 2014 21:27:03 -0700
From: Robert McIntyre <rjmcinty at hotmail.com>
To: "usrp-users at lists.ettus.com" <usrp-users at lists.ettus.com>
Subject: Re: [USRP-users] Success report: B200/USB 3.0 card/Core
	i7-2600
Message-ID: <COL125-W48B1F2791C041AF9C6BEE5D3B80 at phx.gbl>
Content-Type: text/plain; charset="windows-1256"

Ran some tests tonight...

Using:

rx_samples_to_file --rx_rate 32e6 --freq 107.7e6 --time 10 --file ./testoutput --stats

I get either a perfect output, or one or two overflows.  And, that's to either my SSD, or my 7200 RPM SATA-6 drive.  Which I didn't expect, honestly.

But, I tried:

transport_hammer --rx_rate 32e6

And had very bad results.  Not sure what transport_hammer is supposed to do, though.

Anyway, hope this helps!
Robert

To: martin.braun at ettus.com; usrp-users at lists.ettus.com
Date: Tue, 30 Sep 2014 14:23:41 -0700
Subject: Re: [USRP-users] Success report: B200/USB 3.0 card/Core i7-2600
From: usrp-users at lists.ettus.com







Not yet.  I will do that tonight and report back.  The disk, though, will definitely be a concern.  I've got a SSD on a SATA-6G link, but expect it to have problems.



--Robert



From:
Martin Braun via USRP-users

Sent:
?9/?30/?2014 2:06 PM

To:
usrp-users at lists.ettus.com

Subject:
Re: [USRP-users] Success report: B200/USB 3.0 card/Core i7-2600





Hey Robert,



thanks for the data! Out of curiousity, have you also tried streaming to 

disk?



Cheers,

m



On 30.09.2014 12:44, Robert McIntyre via USRP-users wrote:

> Just wanted to drop a quick note about the 100% successful USB transfers

> I'm (finally) able to get with my B200.  I'm using the following hardware:

>

> - Asus P8Z68 motherboard

> <http://www.newegg.com/Product/Product.aspx?Item=N82E16813131790>

> - HooToo HT-PC002 USB 3.0 card

> <http://www.amazon.com/HT-PC001-SuperSpeed-Expansion-Connector-Capacitors/dp/B00B9R59ZE/ref=sr_1_7?ie=UTF8&qid=1412105646&sr=8-7&keywords=hootoo+usb3+card>
 (based

> on VIA VL805)

> - Core i7-2600 cpu

> - 16GB of RAM

> - SSD hard drive

> - OS: Debian Testing

> - UHD: Built from head of master branch

>

> I tried the onboard Intel USB 3 ports, and they weren't able to operate

> at USB 3.0 at all, they only registered the B200 as USB 2.0.  So, I

> added the VL805-based USB 3.0 card, and am able to use the UHD benchmark

> tool with a --rx_rate of 32e6 with zero overruns/underflows, repeatable.

>

> I'll be setting up gnuradio shortly and loading the CPU down a bit and

> see how it works.

>

> I have the same USB 3.0 board in my windows 8.1 box (newer

> CPU/motherboard), and it works well, but still experiences some drops at

> that rate, though 16e6 is rock-solid.

>

> Hope this helps someone else get a good config set up!

>

> Cheers,

> Robert

>

>

>

> _______________________________________________

> USRP-users mailing list

> USRP-users at lists.ettus.com

> http://lists.ettus.com/mailman/listinfo/usrp-users_lists.ettus.com

>






More information about the USRP-users mailing list