[Pw_forum] Use of pool

Huiqun Zhou hqzhou at nju.edu.cn
Mon Mar 2 12:05:06 CET 2009


Liping,

Of course, my tests were made on 8 CPU cores.

Huiqun Zhou
@Nanjing University, China

----- Original Message ----- 
From: "Liping Yu" <lyu7 at ncsu.edu>
To: "PWSCF Forum" <pw_forum at pwscf.org>
Sent: Tuesday, February 24, 2009 9:49 PM
Subject: Re: [Pw_forum] Use of pool


> Dear Huiqun,
> 
> If your results are based on the same number of processes for different
> npools, they make sense to me.   To my understanding, the almost linear
> time scaling of using npool means that the computer time is almost
> linear to npool (and also the total number of procs) only when we fix
> the number of the processors at each pool.
> 
> Please correct me if I am wrong. Thanks!
> 
> Liping Yu
> Department of Physics
> NC State University, USA
> 
> 
> Huiqun Zhou wrote:
>> Dear list users:
>>
>> I happened to test duration times of calculating the system I'm 
>> investigating against number of pools used. There are totally
>> 36 k points. But the results surprised me quite a lot. 
>>
>> no pool:  6m21.02s CPU time,     6m45.88s wall time
>> 2 pools:  7m19.39s CPU time,     7m38.99s wall time
>> 4 pools: 11m59.09s CPU time,    12m14.66s wall time
>> 8 pools: 21m28.77s CPU time,    21m38.71s wall time
>>
>> The machine I'm using is an AMD box with 2 quad core shanghai.
>>
>> Is my understanding of usage of pool wrong?
>>
>> Huiqun Zhou
>> @Nanjing University, China
>> _______________________________________________
>> Pw_forum mailing list
>> Pw_forum at pwscf.org
>> http://www.democritos.it/mailman/listinfo/pw_forum
>>   
> 
> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://www.democritos.it/mailman/listinfo/pw_forum
>



More information about the users mailing list