You do not have permission to delete messages in this group
Copy link
Report message
Show original message
Either email addresses are anonymous for this group or you need the view member email addresses permission to view the original message
to likwid-d...@googlegroups.com
Hello,
I am trying to measure the energy consumed by a particular program by varying the input sizes. I used both likwid-perfctr and likwid-powermeter to do that as they can query the RAPL counters. What I observed was that for large input sizes the energy went down tremendously which was really unexpected. My guess is that there is a limit upto which the msr register can hold values after which it is overflowed. Can you please suggest a workaround for this please.
Thanks
Sonal Aggarwal
unread,
Sep 28, 2014, 9:43:52 AM9/28/14
Reply to author
Sign in to reply to author
Forward
Sign in to forward
Delete
You do not have permission to delete messages in this group
Copy link
Report message
Show original message
Either email addresses are anonymous for this group or you need the view member email addresses permission to view the original message
to likwid-d...@googlegroups.com
I am using the SandyBridge processors with 2 Sockets and 8 cores each.
Thomas Röhl
unread,
Sep 29, 2014, 5:10:03 AM9/29/14
Reply to author
Sign in to reply to author
Forward
Sign in to forward
Delete
You do not have permission to delete messages in this group
Copy link
Report message
Show original message
Either email addresses are anonymous for this group or you need the view member email addresses permission to view the original message
As a workaround you could use the Marker API with regions that require only a small amount of time to reduce the overflow effect. If you want to measure your whole application there is currently no possibility to do it reliably with LIKWID.