[OS X TeX] multi cores

[Alain Schremmer]

On Dec 13, 2009, at 7:52 PM, Peter Dyballa wrote:

>
> Am 13.12.2009 um 23:40 schrieb George Gratzer:
>
>> a long log file displayed in the terminal (would it get faster if  
>> I could suppress that?)
>
> Definitely! "*tex -interaction=batchmode <file>" should save some  
> seconds...
>
>>
>> And at the end:
>>
>> user  	0m1.597s
>
>
> You are using the shell's built-in timer, which cannot be very  
> exact. In bash and GNU's date command, i.e., the command gdate,  
> which has a resolution of nsec, you could run
>
> 	START=$(gdate +%s.%N) ; latex -interaction=batchmode GLT3 ; END=$ 
> (gdate +%s.%N) ; DIFF=$(echo "$END - $START" | bc) ; echo $DIFF
>
> This one-liner will measure from start to end of the execution. The  
> DIFF value times four (you mentioned once that not all the cores  
> are used) might give the "real" time. Another approach is to use  
> Sun's built-in DTraceToolkit:
>
> 	sudo procsystime -eoT latex -interaction=batchmode GLT3
>
> This will deliver some KB of information exact to the nsec (and for  
> me that most time was spent wait'ing 4 something). (Although I  
> really don't know how correct it will be with a multi-core CPU...)  
> In tcsh I use this setting to have some useful and easy to  
> understand (more or less) correct data:
>
> set time=(4 "\
>     Time spent in user mode   (CPU seconds) : %Us\
>     Time spent in kernel mode (CPU seconds) : %Ss\
>     Total time                              : %Es\
>     CPU utilisation (percentage)            : %P")
>
>
> If you want to use dtrace, dtruss, etc. more often it might be  
> useful to set ACLs for your account that allow you to use the  
> dtrace_kernel privileges. (I haven't done it yet.)

I read somewhere that they were getting ready to time the 100 meter  
dash in tenths of milliseconds.

Overwhelmed regards
--schremmer