[gmx-users] Gromacs on Dual Athlons

[Shrivastava, Indira (NCI)]

 Hi, 

 I tried to run a  1ns simulation job on two nodes of a dual Athlon cluster,
in 
LAM and MPI environment. In both the cases, the jobs crashed after 
30-40 ps. I got some spurious (32 digit numbers) angle of rotation values
after a LINCS warning.  

The same job ran through fine on a single node, on 
both LAM and MPI environment. 

Can someone tell me if they have had similar problems ?? 


best wishes 
Indira

--------------------------------------------------------------------
Indira Shrivastava                            		            
LECB,NCI              				
Building 12B, Room B109
MSC 5677    						            
MD 20892 5677 	
 Ph: 301-496-2394
 Fax: 301-402-4724
  e-mail:  shrivasi at mail.nih.gov 
-----------------------------------------------------------------