Hello The difference lies in the P# and L# prefix for these numbers. For historic reason, lstopo shows physical numbers (what you have in /proc/cpuinfo) in graphical mode and logical numbers (in-order numbers generated by hwloc) in textual mode.
You can add -l or -p to force lstopo in logical or physical number mode. See also "Should I use logical or physical/OS indexes? and how?" in https://www.open-mpi.org/projects/hwloc/doc/v1.11.5/a00305.php It causes lots of confusion so I am tempted to change that in the upcoming hwloc 2.0 (or display both by default). Brice Le 02/02/2017 16:19, Gunter, David O a écrit : > Can anyone explain why I get different outputs for Core IDs when using > lstopo’s graphical output versus the text output I get with lstopo —only core? > > On a dual 10-core Sandy Bridge node I see two sets of cores, with IDs > #0,1,2,3,4,8,9,10,11,12. This corresponds with the core IDs I see if I cat > /proc/cpuinfo. > > However, the output from lstopo —only core gives > Core L#0 > Core L#1 > Core L#2 > Core L#3 > Core L#4 > Core L#5 > Core L#6 > Core L#7 > Core L#8 > Core L#9 > Core L#10 > Core L#11 > Core L#12 > Core L#13 > Core L#14 > Core L#15 > Core L#16 > Core L#17 > Core L#18 > Core L#19 > > Why would it be different from the previous? > > Thanks, > david > -- > David Gunter > HPC-ENV: Applications Readiness Team > Los Alamos National Laboratory > > > > > _______________________________________________ > hwloc-users mailing list > hwloc-users@lists.open-mpi.org > https://rfd.newmexicoconsortium.org/mailman/listinfo/hwloc-users _______________________________________________ hwloc-users mailing list hwloc-users@lists.open-mpi.org https://rfd.newmexicoconsortium.org/mailman/listinfo/hwloc-users
