[Nek5000-users] Compilation problem due to large LHIS

[nek5000-users at lists.mcs.anl.gov]

I guess one way to do this is to recycle my g2gi() code. It's fully parallel.
-Stefan

On 1/9/12, nek5000-users at lists.mcs.anl.gov
<nek5000-users at lists.mcs.anl.gov> wrote:
>
> Yes... looking at the code, I see this needs to be done.
>
> Fortunately, it is easy to parallelize.
>
> We'll look into it...
>
> Paul
>
>
> On Mon, 9 Jan 2012, nek5000-users at lists.mcs.anl.gov wrote:
>
>> Hi Paul,
>>
>> We are trying to compute the energy spectrum by transforming the fields
>> onto a particular basis set. The geometry is a box of size 5 x 5 x 1 with
>> 157464 elements. The code to transform onto the basis set takes in
>> hpts.out
>> as the input. The size of the transform is 640 x 640 x 128 which
>> gives 52428800. So you're saying that hpts() has to be parallelized?
>>
>> Thanks,
>>
>> Mani
>> On Jan 9, 2012 9:14 PM, <nek5000-users at lists.mcs.anl.gov> wrote:
>>
>>>
>>> Mani,
>>>
>>> Why do you need it so large?
>>>
>>> Is it possible you can partition this across your processor set?
>>>
>>> Best regards,
>>>
>>> Paul
>>>
>>>
>>> On Mon, 9 Jan 2012,
>>> nek5000-users at lists.mcs.anl.**gov<nek5000-users at lists.mcs.anl.gov>wrote:
>>>
>>>  Dear Nek devs,
>>>>
>>>> I need to put LHIS in the SIZE file to a large number (52428800), but
>>>> with this number I cannot seem to compile despite any amount of
>>>> increasing
>>>> the processors and decreasing LELT. The maximum LP that I tried is 2048
>>>> and
>>>> LELT=77.
>>>>
>>>> I have attached the compile log and the SIZE file. Could I get around
>>>> this problem somehow?
>>>>
>>>> Thanks,
>>>> Mani
>>>>
>>>>  ______________________________**_________________
>>> Nek5000-users mailing list
>>> Nek5000-users at lists.mcs.anl.**gov <Nek5000-users at lists.mcs.anl.gov>
>>> https://lists.mcs.anl.gov/**mailman/listinfo/nek5000-users<https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users>
>>>
>>
> _______________________________________________
> Nek5000-users mailing list
> Nek5000-users at lists.mcs.anl.gov
> https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users
>