Intel® DevCloud
Help for those needing help starting or connecting to the Intel® DevCloud
1617 Discussions

cc1plus: Exhausted virtual memory?

HarshVardhanKumar
New Contributor I
1,293 Views

I'm trying to install a simulator (gem5) for my research. I've been able to install it on a couple of other machines with ubuntu 18.04, however, on the devcloud account, it always gives the error:

cc1plus: out of memory allocating 8224 bytes after a total of 55119872 bytes

I've checked the RAM usage never goes beyond 8.5 GB out of 15GB allocated. So is the case with the swap file.

Please provide any insights into what should be done...

Regards,

Harsh

0 Kudos
1 Solution
AthiraM_Intel
Moderator
1,267 Views

Hi,


Thanks for reaching out to us.


From the screenshot, we could see that you are trying to install from Login node.

Please make sure that you are in compute node for any installation.

You could reach an interactive compute node by using the below command, once you are connected to DevCloud.


qsub -I


For your information, Intel DevCloud® is a shared environment which comes with pre-installed validated Intel® oneAPI software and complimentary packages. As a policy we do not install custom (open source or 3rd party licensed) software to the environment and recommend that you explore available and equivalent oneAPI SW tools or components. You may be able to install software locally in your user folder using “pip install –user”(dash dash user switch allows you to install local in your home directory).


If you face any issue, please let us know.


Thanks.





View solution in original post

2 Replies
AthiraM_Intel
Moderator
1,268 Views

Hi,


Thanks for reaching out to us.


From the screenshot, we could see that you are trying to install from Login node.

Please make sure that you are in compute node for any installation.

You could reach an interactive compute node by using the below command, once you are connected to DevCloud.


qsub -I


For your information, Intel DevCloud® is a shared environment which comes with pre-installed validated Intel® oneAPI software and complimentary packages. As a policy we do not install custom (open source or 3rd party licensed) software to the environment and recommend that you explore available and equivalent oneAPI SW tools or components. You may be able to install software locally in your user folder using “pip install –user”(dash dash user switch allows you to install local in your home directory).


If you face any issue, please let us know.


Thanks.





AthiraM_Intel
Moderator
1,254 Views

Hi,


Thanks for accepting our solution.  If you need any additional information, please submit a new question as this thread will no longer be monitored.


Thanks.


0 Kudos
Reply