[Wien] memory requirements (fwd)
Vicky Tanya Seno
data at csun.edu
Wed May 19 19:35:38 CEST 2004
To Torsten Andersen,
The two machine is questions have the following specs (below). It
seem that different values of NUME and NMATMAX is what is defining the
memory usage of lapw1. But what would you recommend these values of NUME
and NMATMAX be considering the type of system WIEN2k is running on and
the type of job Dr Gotsis is running. How would increasing and
decreasing these values affect the job performance?
Tru64 - DEC Alpha - 4 CPU - 4G Memory
PARAMETER (NMATMAX= 10000)
PARAMETER (NUME= 10000)
IRIX - Origin 3800 - 32 CPU - 32G Memory
PARAMETER (NMATMAX= 100000)
PARAMETER (NUME= 100000)
Tru64
# size -A lapw1
Size of lapw1:914464
Section Size Physical
Address Virtual Address
.dynamic 560
4831839536 4831839536
.liblist 208
4831840096 4831840096
.rel.dyn 32
4831840304 4831840304
.conflic 96
4831840336 4831840336
.dynstr 6048
4831840432 4831840432
.dynsym 11408
4831846480 4831846480
.hash 3968
4831857888 4831857888
.text 379056
4831872000 4831872000
.init 64
4832251056 4832251056
.fini 64
4832251120 4832251120
.got 4224
5368732848 5368732848
.data 22176
5368709120 5368709120
.xdata 1472
5368731376 5368731376
.pdata 1552
4831870448 4831870448
.rdata 80
5368731296 5368731296
.rconst 8592
4831861856 4831861856
.sbss 32
5368737072 5368737072
.bss 470656
5368737104 5368737104
.comment 4176
0 0
#
IRIX
cmt3000 6# size lapw1
Section Size
Physical Virtual
Address Address
.interp 21
268435856 268435856
.MIPS.options 136
268435880 268435880
.dynamic 656
268436016 268436016
.liblist 120
268436672 268436672
.dynstr 4256
268436792 268436792
.hash 3496
268441048 268441048
.dynsym 8640
268444544 268444544
.MIPS.symlib 360
268453184 268453184
.msym 2880
268453544 268453544
.conflict 336
268456424 268456424
.MIPS.stubs 1224
268456760 268456760
.text 262268
268457984 268457984
.init 24
268720256 268720256
.rodata 28616
268795904 268795904
.data 8680
268824520 268824520
.got 2016
268833200 268833200
.lit8 1104
268835216 268835216
.lit4 208
268836320 268836320
.srdata 1158
268836528 268836528
.sdata 4
268837688 268837688
.bss 2768
268837696 268837696
285575 + 40628 + 2768 = 328971
Regards,
Tanya Seno
Harry John Gotsis wrote:
>Dear Tanya;
>
>I'm forwarding the message I've received from Wien
>regarding this big difference in memory
>requirements between 3000 and Alpha.
>
>Let me know your response.
>
>Harry
>
>---------- Forwarded message ----------
>Date: Wed, 19 May 2004 08:09:33 +0200
>From: Torsten Andersen <thor at physik.uni-kl.de>
>Reply-To: wien at zeus.theochem.tuwien.ac.at
>To: wien at zeus.theochem.tuwien.ac.at
>Subject: Re: [Wien] memory requirements
>
>Dear Mr. Gotsis,
>
>I do not think this is reasonable. Maybe you need to check the
>compilers. Does the size of lapw1 change if you change NMATMAX?
>
>If it does, there is a problem with dynamic allocation.
>
>What comes out of the "size lapw1" command?
>
>Best regards,
>Torsten Andersen.
>
>Harry John Gotsis wrote:
>
>
>>Dear Wien users;
>>
>>I'm running a 16 atom cell in two different
>>machines (same parameters):
>>
>>SGI Origin 3000 (Operating System IRIX6.5.20f)
>>DEC/Alpha (Operating System TRU64 v5.1)
>>
>>In the 3000 the memory requirements for lapw1c
>>are 1OG, whereas in the Alpha are 550M.
>>
>>Is this 3 orders of magnitude difference in
>>memory requirements reasonable?
>>
>>Regards,
>>
>>Harry Gotsis
>>
>>_______________________________________________
>>Wien mailing list
>>Wien at zeus.theochem.tuwien.ac.at
>>http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
>>
>>
>>
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://zeus.theochem.tuwien.ac.at/pipermail/wien/attachments/20040519/f2efbd86/attachment.html
More information about the Wien
mailing list