TCE crashes when large number of virtuals is frozen

From NWChem

Viewed 91 times, With a total of 2 Posts
Jump to: navigation, search

Just Got Here
Threads 1
Posts 2
Hi guys,

I've run into a weird problem when running CCSDT calculations with TCE. The calculations crashes if a very large number of virtual orbitals are frozen. So for example, this works for me:

title "1"
memory 14000 mb
start 1
charge 0
geometry units angstrom noautoz noautosym
H -0.00000000 -0.75668992 0.52032190
O 0.00000000 -0.00000000 -0.06557004
H -0.00000000 0.75668992 0.52032190
end

basis spherical
  * library cc-pvtz
end

tce
  CCSDT
  freeze virtual 50
  freeze core
end

task tce energy


Meanwhile, this crashes upon starting the TCE:

title "1"
memory 14000 mb
start 1
charge 0
geometry units angstrom noautoz noautosym
H -0.00000000 -0.75668992 0.52032190
O 0.00000000 -0.00000000 -0.06557004
H -0.00000000 0.75668992 0.52032190
end

basis spherical
  * library cc-pvtz
end

tce
  CCSDT
  freeze virtual 60
  freeze core
end

task tce energy


giving me the following error:

            Memory Information
            ------------------
          Available GA space size is     917500636 doubles
          Available MA space size is     917501036 doubles
 
 Maximum block size        20 doubles
 ------------------------------------------------------------------------
 tce_tile: MA problem                   0
 ------------------------------------------------------------------------
 ------------------------------------------------------------------------


I realize that what I'm trying to do looks weird (or dumb), but I do have a plan! It would be incredibly useful if I could do this type of calculation.

Also: The same thing happens if I run CCSD, either with the TCE or CCSD modules.

What is going on here? What's the difference between freezing 50 and 60 orbitals? Can anyone reproduce this behavior or is it related to my installation? Any help would be much appreciated!

Best regards,

Hannes

Forum Vet
Threads 8
Posts 1388
This system has 53 virtual orbitals. You are asking to freeze 60

Just Got Here
Threads 1
Posts 2
Aha. So it turns out I'm an idiot after all! Sorry about that.

I misread the "number of virtual orbitals" that the TCE prints. I guess that means alpha+beta spinorbitals. I could have figured this out myself...


Forum >> NWChem's corner >> Running NWChem



Who's here now Members 0 Guests 0 Bots/Crawler 1


AWC's: 2.5.10 MediaWiki - Stand Alone Forum Extension
Forum theme style by: AWC