On 15.6.2024. 7:54, Rob Schmersel wrote: > On Fri, 14 Jun 2024 22:20:55 +0200 > Hrvoje Popovski <hrv...@srce.hr> wrote: > >> Hi all, >> >> I have question about cpu output in dmesg. >> I have Fujitsu RX2530m4 with 8 core Intel(R) Xeon(R) Gold 6134 and in >> dmesg I've noticed that core are 0,4,5,7,18,19,21,22 >> >> without HT >> cpu0: smt 0, core 0, package 0 >> cpu1: smt 0, core 4, package 0 >> cpu2: smt 0, core 5, package 0 >> cpu3: smt 0, core 7, package 0 >> cpu4: smt 0, core 18, package 0 >> cpu5: smt 0, core 19, package 0 >> cpu6: smt 0, core 21, package 0 >> cpu7: smt 0, core 22, package 0 >> >> with HT >> cpu0: smt 0, core 0, package 0 >> cpu1: smt 0, core 4, package 0 >> cpu2: smt 0, core 5, package 0 >> cpu3: smt 0, core 7, package 0 >> cpu4: smt 0, core 18, package 0 >> cpu5: smt 0, core 19, package 0 >> cpu6: smt 0, core 21, package 0 >> cpu7: smt 0, core 22, package 0 >> cpu8: smt 1, core 0, package 0 >> cpu9: smt 1, core 4, package 0 >> cpu10: smt 1, core 5, package 0 >> cpu11: smt 1, core 7, package 0 >> cpu12: smt 1, core 18, package 0 >> cpu13: smt 1, core 19, package 0 >> cpu14: smt 1, core 21, package 0 >> cpu15: smt 1, core 22, package 0 >> >> My understanding is that : >> package - cpu socket >> core - physical cpu cores >> smt - core thread >> cpuX - name of core ? >> >> I thought that in my case core should be from 0 to 7 ? >> > <--- snip dmesg ---> > > The Xeon gold 6000 series can have upto 22 cores. The 6134 just has > some of those cores disabled >
Hi, makes sense, tnx ..