Do you have more than one

Chen__Xin · ‎11-13-2014

Hi I use micinfo to check my Phi Card. The the voltage is 0 uv

MicInfo Utility Log
Copyright 2011-2013 Intel Corporation All Rights Reserved.

Created Thu Nov 13 11:03:03 2014

        System Info
                HOST OS                 : Linux
                OS Version              : 2.6.32-279.el6.x86_64
                Driver Version          : 3.2.3-1
                MPSS Version            : 3.2.3
                Host Physical Memory    : 32845 MB

Device No: 0, Device Name: mic0

        Version
                Flash Version            : 2.1.02.0390
                SMC Firmware Version     : 1.16.5078
                SMC Boot Loader Version : 1.8.4326
                uOS Version              : 2.6.38.8+mpss3.2.3
                Device Serial Number     : ADKC32301233

        Board
                Vendor ID                : 0x8086
                Device ID                : 0x225c
                Subsystem ID             : 0x7d95
                Coprocessor Stepping ID : 2
                PCIe Width               : x16
                PCIe Speed               : 5 GT/s
                PCIe Max payload size    : 256 bytes
                PCIe Max read req size   : 512 bytes
                Coprocessor Model        : 0x01
                Coprocessor Model Ext    : 0x00
                Coprocessor Type         : 0x00
                Coprocessor Family       : 0x0b
                Coprocessor Family Ext   : 0x00
                Coprocessor Stepping     : C0
                Board SKU                : C0PRQ-7120 P/A/X/D
                ECC Mode                 : Enabled
                SMC HW Revision          : Product 300W Passive CS

        Cores
                Total No of Active Cores : 61
                ~~Voltage                  : 0 uV~~
                Frequency                : 1238095 kHz

        Thermal
                Fan Speed Control        : N/A
                Fan RPM                  : N/A
                Fan PWM                  : N/A
                Die Temp                 : 38 C

        GDDR
                GDDR Vendor              : Samsung
                GDDR Version             : 0x6
                GDDR Density             : 4096 Mb
                GDDR Size                : 15872 MB
                GDDR Technology          : GDDR5
                GDDR Speed               : 5.500000 GT/s
                GDDR Frequency           : 2750000 kHz
                GDDR Voltage             : 1501000 uV

TaylorIoTKidd · ‎11-13-2014

Do you have more than one board on the host? If so, does it also show 0uV?

Does the coprocessor otherwise perform as it should? Can you log into it? Can you offload?

Regards
--
Taylor

Chen__Xin · ‎11-13-2014

The reason that I recheck the micinfo is that my performace of Phi is slower than that of OpenMP on CPU. only 1/10.

The following the answer of you question:

I have only one card. I therefore don't know the situation of the next card.

I can log on MIC.

I just test intr_sample. It sounds that the card works well.

[root@IC intro_sampleCcp]# ./intro_sampleC.out
Samples started
Checking for Intel(R) Xeon Phi(TM) (Target CPU) devices...

Number of Target devices installed: 1

Offload sections will execute on: Target CPU (offload mode)

PASS Sample01
PASS Sample02
PASS Sample03
PASS Sample04
PASS Sample05
PASS Sample06
PASS Sample07
PASS Sample08
PASS Sample09
PASS Sample10
PASS Sample11
PASS Sample12
PASS Sample13
PASS Sample14

Samples complete

The voltage is a reason that leads to the problem?

Frances_R_Intel · ‎11-13-2014

I found a bug report titled: "micinfo shows "Core Voltage: 0 uV" if run while card is in PC6". (HSD 4868836) I don't know that this is your problem. From what you say, it sounds like you see the 0 voltage regardless of whether the card is executing something (when it would not be in PC6) or not. But just as one thing to try, could you check to see if the power management is enabled on your system? The command 'micctrl --config' will show you your current configuration, including the PowerManagement values. If you see pc6_on, shutdown the coprocessor, change the power management values - the simplest way is probably to use 'micctrl --pm=off' which will disable all the power management - then reboot the coprocessor. If you do this, does the core voltage still show up as 0?

Now, the performance question is another issue and I am given to understand that someone is looking into this for you.

Chen__Xin · ‎11-13-2014

Hi Frances,

Thank you. I have seen the number

        System Info
                HOST OS                 : Linux
                OS Version              : 2.6.32-279.el6.x86_64
                Driver Version          : 3.2.3-1
                MPSS Version            : 3.2.3
                Host Physical Memory    : 32845 MB

Device No: 0, Device Name: mic0

        Version
                Flash Version            : 2.1.02.0390
                SMC Firmware Version     : 1.16.5078
                SMC Boot Loader Version : 1.8.4326
                uOS Version              : 2.6.38.8+mpss3.2.3
                Device Serial Number     : ADKC32301233

        Board
                Vendor ID                : 0x8086
                Device ID                : 0x225c
                Subsystem ID             : 0x7d95
                Coprocessor Stepping ID : 2
                PCIe Width               : x16
                PCIe Speed               : 5 GT/s
                PCIe Max payload size    : 256 bytes
                PCIe Max read req size   : 512 bytes
                Coprocessor Model        : 0x01
                Coprocessor Model Ext    : 0x00
                Coprocessor Type         : 0x00
                Coprocessor Family       : 0x0b
                Coprocessor Family Ext   : 0x00
                Coprocessor Stepping     : C0
                Board SKU                : C0PRQ-7120 P/A/X/D
                ECC Mode                 : Enabled
                SMC HW Revision          : Product 300W Passive CS

        Cores
                Total No of Active Cores : 61
                ~~Voltage                  : 1005000 uV~~
                Frequency                : 1238095 kHz

        Thermal
                Fan Speed Control        : N/A
                Fan RPM                  : N/A
                Fan PWM                  : N/A
                Die Temp                 : 57 C

        GDDR
                GDDR Vendor              : Samsung
                GDDR Version             : 0x6
                GDDR Density             : 4096 Mb
                GDDR Size                : 15872 MB
                GDDR Technology          : GDDR5
                GDDR Speed               : 5.500000 GT/s
                GDDR Frequency           : 2750000 kHz
                GDDR Voltage             : 1501000 uV

Frances_R_Intel · ‎11-14-2014

Good. Unfortunately this is a work around rather than a solution. In a sense, micinfo was right. When the coprocessor is asleep, it isn't drawing any power. However, micinfo does wake the coprocessor up. So, what the question that the bug report is attempting to answer is why, in some cases, does the voltage reading jump in before the card has a chance to wake up. It doesn't do any damage to have micinfo show 0 voltage. So now that we know what is going on, you can enable PC6 if power savings is important to you.

Chen__Xin · ‎11-14-2014

Hi Francse,

Thank you for your response. Of course, I trust the micinfo. However, yesterday, when I double checked my machine, I found the value of Voltage was different between my micinfo and the value listed in Best Practice Guide Intel Xeon Phi V1.1 page 9, in which is value is 1081000 uV. Hence I try to understand the difference.

Why Voltage