AW: NAMD SMP version warning - Reg

From: Norman Geist (norman.geist_at_uni-greifswald.de)
Date: Wed Jul 04 2018 - 09:42:45 CDT

Yes, I know. But I want to see the information please, since NAMD didn’t get the correct information and started everything on only 1 node.

 

Von: Mani Kandan [mailto:manikandand5110_at_gmail.com]
Gesendet: Mittwoch, 4. Juli 2018 12:20
An: Norman Geist <norman.geist_at_uni-greifswald.de>
Betreff: Re: namd-l: NAMD SMP version warning - Reg

 

That has the details of nodes.

 

On Wed, Jul 4, 2018 at 3:48 PM, Norman Geist <norman.geist_at_uni-greifswald.de <mailto:norman.geist_at_uni-greifswald.de> > wrote:

Ok, so what is the content of $LOADL_HOSTFILE

 

Von: Mani Kandan [mailto:manikandand5110_at_gmail.com <mailto:manikandand5110_at_gmail.com> ]
Gesendet: Mittwoch, 4. Juli 2018 12:15
An: Norman Geist <norman.geist_at_uni-greifswald.de <mailto:norman.geist_at_uni-greifswald.de> >
Betreff: Re: namd-l: NAMD SMP version warning - Reg

 

Norman,

 

I want to run over 8 nodes. So I am requesting 128 workers.

 

 

Manikandan D

 

On Wed, Jul 4, 2018 at 3:42 PM, Norman Geist <norman.geist_at_uni-greifswald.de <mailto:norman.geist_at_uni-greifswald.de> > wrote:

If you have 32 cores, why requesting to start 128 workers then?

 

Von: owner-namd-l_at_ks.uiuc.edu <mailto:owner-namd-l_at_ks.uiuc.edu> [mailto:owner-namd-l_at_ks.uiuc.edu <mailto:owner-namd-l_at_ks.uiuc.edu> ] Im Auftrag von Mani Kandan
Gesendet: Mittwoch, 4. Juli 2018 11:58
An: namd-l_at_ks.uiuc.edu <mailto:namd-l_at_ks.uiuc.edu>
Betreff: namd-l: NAMD SMP version warning - Reg

 

Dear all,

 

             I am using NAMD 2.12 SMP (infiniband version), while running in my cluster it shows the following warning

 

Charm++> Running on 1 unique compute nodes (32-way SMP).
Charm++> cpu topology info is gathered in 0.069 seconds.

Charm++> Warning: the number of SMP threads (136) is greater than the number of physical cores (32), so threads will sleep while idling. Use +CmiSpinOnIdle or +CmiSleepOnIdle to control this directly.

 

 

                  How can i rectify this problem? and performance also very slow. I am using following command to run namd: charmrun ++mpiexec +p128 ++ppn 16 ++nodelist $LOADL_HOSTFILE namd2 <file>

I am having 16 cores per processor.

 

Kindly help me in these regard.

 

 

Manikandan D

 

 

This archive was generated by hypermail 2.1.6 : Mon Dec 31 2018 - 23:21:14 CST