PCM reporting 1 socket on 2 socket Haswell system

[UL_HPC_Sysadmins]

Dear all,

I'm testing Intel PCM (v2.10) on a cluster (Debian 7, 3.2 kernel) with Intel Haswell and SandyBridge CPUs (2 sockets per system). The pcm-power.x tool reports only 1 socket for the Haswell systems, while for the SandyBridge ones the correct 2 sockets. Anyone else experience this issue?

Wrong output on Haswell nodes:

$ pcm-power.x
 Intel(r) Performance Counter Monitor V2.10 (2015-11-17 09:01:38 +0100 ID=cd66c34)

 Power Monitoring Utility
 Copyright (c) 2009-2015 Intel Corporation
Number of physical cores: 1
Number of logical cores: 24
Number of online logical cores: 24
Threads (logical cores) per physical core: 24
Num sockets: 1
Physical cores per socket: 1
Core PMU (perfmon) version: 3
Number of core PMU generic (programmable) counters: 8
Width of generic (programmable) counters: 48 bits
Number of core PMU fixed counters: 3
Width of fixed counters: 48 bits
Nominal core frequency: 2500000000 Hz
Package thermal spec power: 120 Watt; Package minimum power: 61 Watt; Package maximum power: 240 Watt; 
PCM Warning: the bus for socket 0 on system with 1 sockets could not find via PCI bus scan. Using cpubusno register. Bus = 127
2 memory controllers detected with total number of 4 channels. 
[...]

$ cat /proc/cpuinfo  | egrep 'model name|physical id' | sort | uniq -c
     24 model name      : Intel(R) Xeon(R) CPU E5-2680 v3 @ 2.50GHz
     12 physical id     : 0
     12 physical id     : 1

Good output on SandyBridge nodes:

$ pcm-power.x 

 Intel(r) Performance Counter Monitor V2.10 (2015-11-17 09:01:38 +0100 ID=cd66c34)

 Power Monitoring Utility
 Copyright (c) 2009-2015 Intel Corporation
Number of physical cores: 16
Number of logical cores: 16
Number of online logical cores: 16
Threads (logical cores) per physical core: 1
Num sockets: 2
Physical cores per socket: 8
Core PMU (perfmon) version: 3
Number of core PMU generic (programmable) counters: 8
Width of generic (programmable) counters: 48 bits
Number of core PMU fixed counters: 3
Width of fixed counters: 48 bits
Nominal core frequency: 2200000000 Hz
Package thermal spec power: 95 Watt; Package minimum power: 52 Watt; Package maximum power: 150 Watt; 
ERROR: QPI LL monitoring device (0:63:8:2) is missing. The QPI statistics will be incomplete or missing.
ERROR: QPI LL monitoring device (0:63:9:2) is missing. The QPI statistics will be incomplete or missing.
Socket 0: 1 memory controllers detected with total number of 4 channels. 0 QPI ports detected.
ERROR: QPI LL monitoring device (0:127:8:2) is missing. The QPI statistics will be incomplete or missing.
ERROR: QPI LL monitoring device (0:127:9:2) is missing. The QPI statistics will be incomplete or missing.
Socket 1: 1 memory controllers detected with total number of 4 channels. 0 QPI ports detected.
[...]
$ cat /proc/cpuinfo  | egrep 'model name|physical id' | sort | uniq -c
     16 model name      : Intel(R) Xeon(R) CPU E5-2660 0 @ 2.20GHz
      8 physical id     : 0
      8 physical id     : 1

Thank you for your time,
Best regards,
Valentin Plugaru
--
Valentin Plugaru | valentin.plugaru@uni.lu
Parallel Computing and Optimization Group
CSC Research Unit, University of Luxembourg
PGP KeyID 0x324FADC2

[TimP]

This appears more topical for the companion forum on tuning and platform monitoring..

[UL_HPC_Sysadmins]

Indeed, at a quick look I had seen another post on PCM here and not in the Platform Monitoring section. I'm not able to move the topic there though.

[Roman_D_Intel]

The /proc/cpuinfo from the Haswell system indicates that there are only 24 logical cores on the system which corresponds to just a single socket with Intel Xeon E5 2680 v3 processor.

I think that the Cluster on Die feature is enabled on this system resulting in two package ids per socket visible in Linux.

It is also worth trying the latest Intel PCM Version 2.11

Thanks,

Roman