[linrad] Re: From the top down (inexpensively)

[<leif@xxxxxxxxxx>]

Hi Zaba,


> As a matter of
> fresh curiosity I wonder, whether multi-threaded 'Linrad' will 
> see daylight
> soon, and whether it will run in the lir02-series (or higher up 
> the ladder)?
I have a precursor of 2.00 running. By now it is just a multi-
threaded version for the main processing, setup, calibration and
everything else is on a single thread. It is still very buggy,
I just learned that the proper way to make an error exit is the 
one I already implemented. This was bad news, it makes the transfer
from single threaded more difficult for all the support functions
with their myriad of error exits. Seems I have to add one extra thread
with the only task to kill everything else.

Now, multi-threading in itself does not give any advantage, actually
it makes the program less efficient and more memory hungry. Not very
big losses, but on the MMX266 MHz it is the difference between 
working fine at 97% CPU load and not working.......

With less demanding parameters I see an increase of the CPU load by
about 5% (from 90 to 95%) so it is not a big problem and the cost
of the overhead should be about 1% with normal wideband parameters
that require about 5 times more computing. An increase from 25%
to 25.25% is quite acceptable I think:-)

I do not yet know whether multi-threading makes operation impossible
on 486 machines. The 166 MHz Pentium 1 works fine but I have not looked 
at the timing - for 3kHz bandwidth the CPU has a lot of spare time:-)

73

Leif / SM5BSZ


 


#############################################################
This message is sent to you because you are subscribed to
  the mailing list <linrad@xxxxxxxxxxxxxxxxxxxxx>.
To unsubscribe, E-mail to: <linrad-off@xxxxxxxxxxxxxxxxxxxxx>
To switch to the DIGEST mode, E-mail to <linrad-digest@xxxxxxxxxxxxxxxxxxxxx>
To switch to the INDEX mode, E-mail to <linrad-index@xxxxxxxxxxxxxxxxxxxxx>
Send administrative queries to  <linrad-request@xxxxxxxxxxxxxxxxxxxxx>

LINRADDARNIL