Re: [Discuss-gnuradio] Speed test went wrong.

[James Shimer]

The throttle block has some peculiar limitations.  For example, if you try to set the throttle very low (like < 1k) you'll find that there is minimum threshold and what will happen is it will still give you way more samples than you expect.  Before I found the head block, which allows you to terminate the FG after a certain number of samples, I tried using the throttle to slow things way down and just kill the FG manually.  I was using the file sink to log my data and I saw way too much data.  I looked at the throttle block code, it looked like there were some conversion between time formats that would loose precision.  It wouldn't be a surprise to me if there were the throttle block behaved the same way for very high > 1M rates (that is just ignore what you asked for and perform some sort of max threshold behavior).   I would suggest just retesting with out the throttle block, since you're looking to see what you're machine can handle anyway....

Sent from Outlook

---

From: Discuss-gnuradio <discuss-gnuradio-bounces+address@hidden> on behalf of Booth <address@hidden>
Sent: Tuesday, April 25, 2017 6:18 AM
To: address@hidden
Subject: [Discuss-gnuradio] Speed test went wrong.

 

Dear all,
 I wondered what would be the maximum sample rate that my PC can handle for
a basic psk mod-demod. I have used the “prbs_test.grc” from gr-mapper as it
contains all the necessary blocks. I changed the sample rate for the
throttle to 32k, 320k,10M, 100M even 10G but the system did not break down.
I repeated the experiment using different examples but the result was the
same.
What went wrong and do you know any appropriate projects for speed
measurements.



--
View this message in context: http://gnuradio.4.n7.nabble.com/Speed-test-went-wrong-tp63632.html
Sent from the GnuRadio mailing list archive at Nabble.com.

_______________________________________________
Discuss-gnuradio mailing list
address@hidden
https://lists.gnu.org/mailman/listinfo/discuss-gnuradio