[Nek5000-users] ?= Run nek5000 in BlueGene supercalculator (IBM=?utf-8?q? compiler)

nek5000-users at lists.mcs.anl.gov nek5000-users at lists.mcs.anl.gov
Fri Mar 24 03:20:34 CDT 2017


If there is indeed a bug with load_fld please report an issue on Github an provide the necessary details. This way we can fix it.

Stefan 

On 24 Mar 2017, at 09:16, "nek5000-users at lists.mcs.anl.gov <mailto:nek5000-users at lists.mcs.anl.gov> " <nek5000-users at lists.mcs.anl.gov <mailto:nek5000-users at lists.mcs.anl.gov> > wrote:

Hello again,

  1. I use "mpixlf77" and "mpixlc_r" for fortran and C compilers, 
respectively.

  2. Is it possible for you to run the case that you are attempting on 
IDRIS, on
a different server? Perhaps, by comparing the output files for the two you
might get an idea of resolving the problem. Can you check if you are 
able to
run examples from NeK repository on IDRIS?

  3. I would suggest that you try load_fld subroutine first. It might 
work for you
as we used "load_fld" to read base flow as well perturbation variables.
To circumvent this problem on IDRIS, we use "mfi" subroutine of nek with 
slight
modifications.

Cheers.

N.

On Friday 24 March 2017 09:00 AM, nek5000-users at lists.mcs.anl.gov <mailto:nek5000-users at lists.mcs.anl.gov> wrote:
>   Hello,
>
> Thank you very much for your answear, could you please share more detail about your changes?
>
> 1: What compilers did you chose?
> 2: I will do that
> 3: I read only standard file, I don't have to read fld (yet), the fileds is initialized with an analytical form, and it fail just after the setup phrase, it show NaN in one columm of Hemholtz step.
> But I will have to read on fld file, what alternative methode did you use?
>
> Thank you again,
>
> Best,
>
> Le Vendredi 24 Mars 2017 08:25 CET, nek5000-users at lists.mcs.anl.gov <mailto:nek5000-users at lists.mcs.anl.gov> a écrit:
>   
>> Hello,
>>
>>    I recently moved my computations to the blue-gene server of
>> IDRIS. I had to make the following modifications to make the code
>> run on this server:
>>
>> 1. Change the compiler settings. A page containing the details
>> of the various compilers are given on IDRIS website.
>>
>> 2. There were some issues with the sub-block submission. Sometimes the
>> the computations hung displaying "permission denied". So I was advised
>> to use compute blocks in multiple of 64.
>>
>> 3. Can you check which step of your code gives NaN? Is it because of
>> an error in reading any file? In my simulations, apparently, "load_fld'
>> subroutine was causing a problem so I used an alternate method to read
>> the files.
>>
>> Hope these help.
>>
>> N.
>>
>> On Thursday 23 March 2017 08:32 PM, nek5000-users at lists.mcs.anl.gov <mailto:nek5000-users at lists.mcs.anl.gov> wrote:
>>> Hi,
>>>
>>> I have trouble running Nek in BlueGene in IDRIS, French national supercalculator: the compilation show no error, but the output give NaN before the time loop start. I looked in the internet and I found 1 person ask almost the same question 4 years ago (In my case the compilation was fine), but I can't find the answear.
>>>
>>> Are there anyone ran Nek5000 in IDRIS can share their experience?
>>>
>>> Thank you guy,
>>>
>>>
>>>
>>> _______________________________________________
>>> Nek5000-users mailing list
>>> Nek5000-users at lists.mcs.anl.gov <mailto:Nek5000-users at lists.mcs.anl.gov> 
>>> https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users
>> _______________________________________________
>> Nek5000-users mailing list
>> Nek5000-users at lists.mcs.anl.gov <mailto:Nek5000-users at lists.mcs.anl.gov> 
>> https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users
>   
>   
>   
>   
>
>
> _______________________________________________
> Nek5000-users mailing list
> Nek5000-users at lists.mcs.anl.gov <mailto:Nek5000-users at lists.mcs.anl.gov> 
> https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users
>

_______________________________________________
Nek5000-users mailing list
Nek5000-users at lists.mcs.anl.gov <mailto:Nek5000-users at lists.mcs.anl.gov> 
https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mcs.anl.gov/pipermail/nek5000-users/attachments/20170324/d5da8145/attachment.html>


More information about the Nek5000-users mailing list