******************* This email originates from outside Imperial. Do not click on links and attachments unless you recognise the sender. If you trust the sender, add them to your safe senders list https://spam.ic.ac.uk/SpamConsole/Senders.aspx to disable email stamping for this address. ******************* Hi, I am running Nektar++ compressible code and I ssh in one of nodes. It looks like the cores are not all running at ~100% in the CPU column. Not sure if other people had this problem in the past but it looks like that this could affect the parallel efficiency. I don't have access to a profiler to check where the penalty comes from but I was thinking to check with the community. Cheers, Andrei [cid:image002.jpg@01D7005F.F339E5F0] Not Subject to Export Control This information is provided by Rolls-Royce in good faith based upon the latest information available to it; no warranty or representation is given; no contractual or other binding commitment is implied. (c) 2021 Rolls-Royce The data contained in, or attached to, this e-mail, may contain confidential information. If you have received it in error you should notify the sender immediately by reply e-mail, delete the message from your system and contact +44 (0) 3301235850 (Security Operations Centre) if you need assistance. Please do not copy it for any purpose, or disclose its contents to any other person. An e-mail response to this address may be subject to interception or monitoring for operational reasons or for lawful business practices. (c) 2021 Rolls-Royce plc Registered office: Kings Place, 3rd Floor 90 York Way, London, N1 9FX Company number: 1003142. Registered in England
Hi Andrei, Not entirely sure what may be going on here but you may need to bind the process to CPUs. It the node is also open to other users there can be contention or is this a single user parallel facility? Finally another thing to check might be whether it runs at 100% with fewer nodes in case you are at a strong scaling limit for this run? Best regards, Spencer. On 11 Feb 2021, at 10:23, Cimpoeru, Ionut Andrei <IonutAndrei.Cimpoeru@Rolls-Royce.com<mailto:IonutAndrei.Cimpoeru@Rolls-Royce.com>> wrote: This email from IonutAndrei.Cimpoeru@Rolls-Royce.com<mailto:IonutAndrei.Cimpoeru@Rolls-Royce.com> originates from outside Imperial. Do not click on links and attachments unless you recognise the sender. If you trust the sender, add them to your safe senders list<https://spam.ic.ac.uk/SpamConsole/Senders.aspx> to disable email stamping for this address. Hi, I am running Nektar++ compressible code and I ssh in one of nodes. It looks like the cores are not all running at ~100% in the CPU column. Not sure if other people had this problem in the past but it looks like that this could affect the parallel efficiency. I don’t have access to a profiler to check where the penalty comes from but I was thinking to check with the community. Cheers, Andrei <image002.jpg> Not Subject to Export Control This information is provided by Rolls-Royce in good faith based upon the latest information available to it; no warranty or representation is given; no contractual or other binding commitment is implied. © 2021 Rolls-Royce The data contained in, or attached to, this e-mail, may contain confidential information. If you have received it in error you should notify the sender immediately by reply e-mail, delete the message from your system and contact +44 (0) 3301235850 (Security Operations Centre) if you need assistance. Please do not copy it for any purpose, or disclose its contents to any other person. An e-mail response to this address may be subject to interception or monitoring for operational reasons or for lawful business practices. (c) 2021 Rolls-Royce plc Registered office: Kings Place, 3rd Floor 90 York Way, London, N1 9FX Company number: 1003142. Registered in England _______________________________________________ Nektar-users mailing list Nektar-users@imperial.ac.uk<mailto:Nektar-users@imperial.ac.uk> https://mailman.ic.ac.uk/mailman/listinfo/nektar-users
participants (2)
-
Cimpoeru, Ionut Andrei
-
Sherwin, Spencer J