Servidor

proc_bind

I specify for example my env variables as follows: "mic_kmp_place_threads=Xc,1t", "mic_omp_num_threads=X", "mic_kmp_affinity=scatter". I keep X smaller than available cores on mic.

1) Would there be any difference if I set my "mic_omp_proc_bind" to spread or close?

2) Also are there any documentation on "mic_omp_proc_bind"?

mpss 3.4 - No Mic cards found or specified on command line

I bet this is most common issue. We have installed the Card and the MPSS software. We are getting this message.

# micctrl --initdefaults
[Warning] No Mic cards found or specified on command line

We are able to detect the module:

# find /lib/modules -name mic.ko
/lib/modules/3.10.0-123.el7.x86_64/extra/mic.ko

OS sees the card:

 lspci | grep coprocessor

04:00.0 Co-processor: Intel Corporation Xeon Phi coprocessor SE10/7120 series (rev 11)

And here are other details:

using version 2015 of the compiler to build for mic does not seem to work on RHEL 6

I am trying to compile for both MIC and the host, and am hitting this error: x86_64-k1om-linux-ld: /lib64/libc.so.6: version `GLIBC_2.14' not found (required by x86_64-k1om-linux-ld) x86_64-k1om-linux-ld: /lib64/libc.so.6: version `GLIBC_2.14' not found (required by /opt/mpss/3.4/sysroots/x86_64-mpsssdk-linux/usr/lib/k1om-mpss-linux/../libz.so.1) GLIBC on RHEL 6 is 2.12.x.

Compiler version and offloading errors

while playing with the offloading examples found on this forum I ran into a little snag:

host A: Scientific Linux 6.5, Intel compiler v14, Xeon Phi 7100, mpss 3.3.2 stack

host B: CentOS 6.5, Intel compiler v15, 2 x Xeon Phi 5100, mpss 3.3.2 stack

When I compile an offload example on host A, then copy it over to host B and I try to run it I get:

[hostB] $ ./offload1.icc14
offload error: target executable is not available

It does not work even if I copy over the offload libraries from icc v14 to the new environment:

Compiler version and offloading errors

while playing with the offloading examples found on this forum I ran into a little snag:

host A: Scientific Linux 6.5, Intel compiler v14, Xeon Phi 7100, mpss 3.3.2 stack

host B: CentOS 6.5, Intel compiler v15, 2 x Xeon Phi 5100, mpss 3.3.2 stack

When I compile an offload example on host A, then copy it over to host B and I try to run it I get:

[hostB] $ ./offload1.icc14 
offload error: target executable is not available

It does not work even if I copy over the offload libraries from icc v14 to the new environment:

Is there any document telling the latency of every instruction?

hi, buddies

I am writing a Phi native application, but the utilization of CPU is very low (about 25%) and the CPI is a little high (about 3). I used plenty vector instruction in the application, and suspect that's the reason for the bad performance. I wonder if there is any official document or manual telling the latency of every instruction of Xeon Phi, I have searched but only got some empirical experience like this

http://arxiv.org/abs/1310.5842

so, please let me know if you have ideas, thanks.

Suscribirse a Servidor