Newsgroups: comp.parallel
From: Mr P Anderson <pda@doc.ntu.ac.uk>
Subject: Re: Problems with use of TMS320C40's
Organization: Nottingham Trent University
Date: 12 Sep 1995 13:28:10 GMT
Message-ID: <4341ta$9ps@usenet.srv.cis.pitt.edu>

esrlk@eng.warwick.ac.uk (A L Roberts) wrote:

[stuff deleted]

>Using the the DB40 Debugger it appears that running the program on one
>processor the computation time to calculate one value takes 46 clock
>cycles (this time does not include file operations, declaring
>variables, etc). If the program is then split onto two processors the
>computation time on each processor reduces to 27 clock cycles per
>value, but the transfer of data between the two processors takes 192
>clock cycles per value (using the chan_in_message and chan_out_message
>commands does this seem reasonable ?).  The time taken to calculate
>all the necessary points actually increases when  run on two or more
>processors. 

How many bits in your value that is transferred? I don't use 3L so I have
no idea whether these transfer times are reasonable for their libraries.
However the C40 comm ports transfer at 160 Mega bits a second. This gives
a bit transfer at around 6ns (correct me if my maths are gone wrong here).
In your 192 cycles, with CPU at 40MHz, you should be able to transfer
(in theory) 1280 bits. No doubt the calling of chan_out etc will have some
overhead.

-- 
Paul Anderson,                   |\      _,,,---,,_       
Software Engineer,          zzz. /,`.-'`'    -.  ;-;;,_. 
Computing Dept.,                |,4-  ) )-,_. ,\ (  `'-' 
Nottingham Trent Uni.          '---''(_/--'  `-'\_)

